Modelling the role of epiphyton and water level for submerged macrophyte development with a modified submerged aquatic vegetation model in a shallow reservoir in China

Ecological Engineering 81 (2015) 123–132

Contents lists available at ScienceDirect

Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng

Modelling the role of epiphyton and water level for submerged macrophyte development with a modified submerged aquatic vegetation model in a shallow reservoir in China Chen Zhang a, * , Xueping Gao a , Liyi Wang b , Xiaojun Chen b a b

State Key Laboratory of Hydraulic Engineering Simulation and Safety, Tianjin University, Tianjin 300072, China Yuqiao Reservoir Administrative Bureau of Luan River-Tianjin Water Diversion Project, Tianjin 301900, China

A R T I C L E I N F O

A B S T R A C T

Article history: Received 22 August 2014 Received in revised form 19 January 2015 Accepted 5 April 2015 Available online xxx

Light intensity plays an important role in determining the distribution of submerged macrophyte. A modified SAVM (M-SAVM) was constructed to simulate the role of epiphyton and water level for the biomass and distribution of Potamogeton crispus in the Yuqiao Reservoir in China. M-SAVM is developed by modification of the light attenuation equation, which is determined by the water transparency (Secchi depth) and epiphyton. The model was calibrated and verified by biomass using two datasets, from the seedling establishment until dying out, in 2008 and 2009. Five hydraulic scenarios were simulated by MSAVM to analyze the relationship between biomass and water depth. Results showed that epiphyton increase had a slightly low light intensity limitation coefficient to suppress plant growth in M-SAVM. Significant negative correlation (p < 0.01, r =
0.97) between biomass and water depth existed in the reservoir. The biomass increases under low water levels due to increasing underwater light intensity and decreases when the water level is raised. M-SAVM could be a useful tool for submerged macrophyte management in the reservoir and for maintaining intermediate vegetation biomass by fluctuating water level strategies in shallow lakes. ã 2015 Elsevier B.V. All rights reserved.

Keywords: Submerged macrophytes Modified SAVM Light attenuation Epiphyton Water-level fluctuations Yuqiao Reservoir

1. Introduction Shallow lakes have been the subject of intensive research on eutrophication and extensive efforts to limit phytoplankton production. They are typically in one of two self-stabilizing equilibrium states: a ‘clear’ state with submerged macrophytes or a ‘turbid’ state dominated by phytoplankton (Scheffer et al., 1993b; Janse et al., 2008, 2010). Submerged macrophyte and phytoplankton components of eutrophic, shallow lakes frequently undergo dynamic changes in composition and abundance with important consequences for lake functioning and stability (Sayer et al., 2010; Spoljar et al., 2012). Eutrophication is likely a major cause of reductions in submerged macrophyte (Jeppesen et al., 1997; Boesch et al., 2001). Aquatic vegetation sometimes plays a significant role in the ecological functioning of lake ecosystems. Submerged macrophyte is considered an important tool for lake restoration. Previous studies suggest that nutrient levels are generally lower in

* Corresponding author. Tel.: +86 22 27404178; fax: +86 22 27890287. E-mail addresses: [email protected] (C. Zhang), [email protected] (X. Gao), [email protected] (L. Wang), [email protected] (X. Chen). http://dx.doi.org/10.1016/j.ecoleng.2015.04.048 0925-8574/ ã 2015 Elsevier B.V. All rights reserved.

submerged macrophyte dominated waterbodies (e.g., Li et al., 2007; Kosten et al., 2009; Moreno, 2011). Watershed-derived nutrient loading has also caused an increase in algal biomass and a degradation and loss of macrophyte habitat. Examples are the soft water lakes of northern Europe e.g., Lake Ladoga, Russia (Murphy, 2002) as well as the estuarine system e.g., Chesapeake Bay USA (Boesch et al., 2001) and Waquoit Bay USA (Deegan et al., 2002). In general, the main benefit of abundant submerged macrophytes is that vegetation seems to be the obvious solution to restore shallow lakes by phosphorus and nitrogen removal. The submerged macrophyte dominated lake and river systems in Florida (Knight et al., 2003; Dierberg et al., 2005), Lake Chiemsee and Lake Starnberg in Germany (Melzer, 1999), and freshwater wetland of the Upper Cooper River Estuary in South Carolina (McKellar et al., 2007) are some examples. Furthermore, submerged macrophytes of lakes are valuable as fisheries economically and for recreational tourism. Enhanced submerged macrophytes under management strategies would increase fishery production (Ma et al., 2010). Restoration work, such as reinforcing the lakes’ shoreline vegetation, has been achieved in shallow Dutch lakes (Gulati and Van Donk, 2002). However, dense beds of aquatic macrophytes are often a nuisance to boaters and swimmers and may obstruct water flow

124

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

(Van Nes et al., 1999; Wang et al., 2006). Rooted submerged macrophytes can be a nuisance for surface water systems, affecting the channel hydraulics by restricting the conveyance volume, they also affect the water quality by plant cycling of nutrients and inorganic carbon and the impacts of biomass after sloughing (Berger and Wells, 2008). Therefore, aiming for an intermediate level of vegetation biomass seems to be a good initial solution for this controversy. Submerged macrophyte growth is limited by light, temperature, nutrients, and other factors such as salinity and self-shading. Epiphyton also plays an important role in submerged macrophyte development. In previous research, the relationships among the nutrients, submerged macrophytes, epiphyton, phytoplankton, and grazing invertebrates were studied in shallow lakes (Phillips et al., 1978; Brönmark & Weisner 1992; Jones et al., 1998, 2002; Jones and Sayer, 2003; Beresford and Jones, 2010). Increased epiphyton loads are detrimental to the plant growth through competition for light and carbon dioxide (Sand-Jensen, 1977; Jones et al., 2002). Light intensity is identified as a primary controlling factor for submerged macrophyte growth. The deeper into water that the light can penetrate, the deeper the depths at which photosynthesis can occur. Depths depend on water-level fluctuations (WLFs) may strongly affect light available for submerged macrophyte in shallow lakes. Managed WLFs have become an important and useful tool for lake management in European lakes since the end of the last century (e.g., Ter Heerdt and Drost, 1994; Coops and Hosper, 2002; Naselli-Flores and Barone, 2005; Hilt et al., 2006; Leira and Cantonati, 2008; Paillisson and Marion, 2011). WLFs may have a strong impact on sediment and nutrient fluxes in shallow lakes, mainly through the development of vegetation covering both the shorelines and the lake bottoms (Coops and Hosper, 2002). Consequently, understanding the role of WLFs in ecosystem functioning has become even more crucial e.g., temporal and

spatial scales, biotic responses, regional differences, influence of climate and climate change (Coops et al., 2003). For almost 30 years, mathematical models have guided management efforts to reduce eutrophication in waterbodies. Many eutrophication models are well-established e.g., CE-QUAL-W2 (Cole and Buchak,1995; Berger and Wells, 2008), CE-QUAL-ICM (Cerco and Cole, 1993, 1994), EFDC (Hamrick, 1992), WASP (Wool et al., 2002), and MIKE 3 (DHI, 2001). Complex ecosystem models with a macrophyte component have already been developed, such as Megaplant (Scheffer et al., 1993a), SAGA (Hootsmans, 1994, 1999), PCLake (Janse, 1997, 2005; Janse et al., 2008, 2010), SAVM (Cerco and Moore, 2001; Cerco et al., 2002; Jin et al., 2007; Jin and Ji, 2013), and Charisma (Van Nes et al., 2002, 2003). All of these models have light attenuation equations available; nevertheless, the empirical coefficient (background light attenuation coefficient) plays a critical role in the equations. Thus, the light attenuation equation is tried to modify, which is determined by the water transparency and epiphyton – a modified SAVM (M-SAVM). The objectives of this paper are: (1) to utilize M-SAVM to analyze the role of epiphyton for light attenuation and submerged macrophyte development in the Yuqiao Reservoir, (2) to illustrate the quantitative relationship between biomass and water depth. This study would provide information for the management of submerged macrophytes in a shallow reservoir or a shallow lake using hydraulic control strategies. 2. Material and methods 2.1. Study area The Yuqiao Reservoir (Fig. 1) is located in Tianjin, China, and has been the only water supply source in Tianjin since 1983. This shallow reservoir (Zmax 12.16 m, Zmean 4.74 m) has a watershed area of 2060 km2, storage capacity of 0.385 billion m3 and surface area

Fig. 1. Yuqiao Reservoir (117 340 E, 40 020 N) watershed and the monitoring stations (Stations S1–S7).

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

of 86.8 km2 at an elevation of 21.16 m in accordance with the Dagu elevation standard (China). The Lin-Sha and Li Rivers are tributaries of flow into the reservoir, Zhou culvert is the outflow of the reservoir. There were 0.44 billion m3 and 0.53 billion m3 water diverted from Li River to the Yuqiao Reservoir in 2008 and 2009, respectively. The annual WLFs ranged from 2.63 m to 3.00 m during the period of 2008–2009. Seven monitoring sites are located in the Yuqiao Reservoir (Fig. 1). Stations S1–S3 have been used to collect water quality parameters since 1989, which typifies the water in the uppermost part of the reservoir (near the tributaries), inside the reservoir and the water output from the reservoir, respectively. Because the reservoir is so broad, Stations S4–S6 have been additionally investigated since 1999 (Chen et al., 2012). Station S7 has been investigated since 2008 due to abundant submerged macrophyte. Survey on the reservoir habitats was conducted from May 2006 to October 2007, according to the Standard for the investigation of reservoir fishery resources of China (Ministry of water resources of the People’s Republic of China, 1996). Macrophyte communities in the region were mostly composed of six emergent macrophytes (e.g., Phragmites communis Trin., Polygonum hydropiper Linn., etc.), 13 submerged macrophytes (e.g., Potamogeton crispus Linn., Myriophyllum spicatum Linn., Potamogeton maackianus A. Benn., etc.), and two floating macrophytes (Trapa incise Sieb. et Zucc. and Nymphoides peltata (Gmel) O. Kuntze.). The reservoir was previously dominated by P. crispus with 35.0–61.5% coverage and a mixture of floating macrophytes (8.3% coverage) in May 2006. The wet mass of P. crispus was approximately 2.8–4.6 kg m
2 (Zhang et al., 2011). The first shoot of P. crispus arrives approximately in late March, and they all have nearly died out by the end of June or early July (Zhang et al., 2011). During the growth period of P. crispus, maximum algal biomass reproduces with less than 5 mg l
1 chlorophyll-a. Little epiphyton was found on submerged macrophyte leaf surfaces. The reservoir is a phosphorus-limited environment (Chen et al., 2012), the total phosphorus (TP) concentrations have an annual value of 0.027  0.002 mg l
1 during 2008–2009 at Station S2. 2.2. The databases and biomass estimation method Considering the first shoot and complete decay period of P. crispus, M-SAVM was applied to an almost 5-month period (March 1–July 18) in 2008 and 2009. Monitoring stations S1–S7 and sampling for water quality parameters, observed hydrodynamic and water quality data, and meteorological data have been previously described (Chen et al., 2012; Zhang et al., 2013). The Secchi depth has been monitored on a semi-monthly basis at Stations S1–S7. The biomass has been investigated on a daily to weekly basis during the growth period of P. crispus at Stations S2 and S4–S7 since 2008. In addition, water quality parameters including the Secchi depth have been monitored on a daily to weekly basis during this period at Stations S1–S7. P. crispus dry biomass (Bd , g DM m
2) was estimated by first determining the average biomass per individual stem (Bi ) at each sampled location and then multiplying this value by the number of stems per square meter (Ns ) (Piedade et al., 1991). Individual stem biomass was obtained by the subsampling method (Eq. (1)) (Silva et al., 2010). In general, the total organic carbon (TOC) content is 30–40% in the submerged macrophyte (dry mass). TOC was assumed a slightly higher value of 50%, considering the loss of biomass during the sampling and measuring process. Bd ¼ Bi  Ns ¼

Bq  Ns Iq

(1)

where Iq is the number of stems harvested for biomass in the partial sample in each quadrat, and Bq is the measured biomass of

125

the partial sample in each quadrat. To harvest the plant, the water region at each station was divided into 12 sub-regions (a rectangle of 3 m by 4 m, each sub-region with an area of 1 m2). Plants were harvested sequentially in each sub-region with a quadrat (0.2 m radius sickle) at Stations S2 and S4–S7 on dates of sampling. The epiphytic chlorophyll-a was measured according to the method of Qin et al. (2006). During the sampling for plants, epiphyton was measured at Station S2 in 2009. 2.3. The governing biomass-balance equations of the SAVM The state variables and basic principles of the SAVM were based on Wetzel and Neckles (1986) and Madden and Kemp (1996). The model incorporates three state variables: shoots (biomass in the water column), roots (biomass in the sediment bed), and epiphyton (attached growth). The kinetic mass balance equations for shoot, root, and epiphyton may be expressed as Eqs. (2)–(4) (Cerco and Moore, 2001; Ji, 2008).

@Bs ¼ @t



1
f Pr Ps
Rs
Ls Bs þ Jrs

(2)

@Br ¼ f Pr Ps Bs
ðRr þ Lr ÞBr
Jrs @t

(3)

@Be ¼ ðPe
Re
Le ÞBe @t

(4)

where Bs , Br , and Be are the shoot, root, and epiphyton biomass (g C m
2), respectively; Ps is the shoot production (day
1); f Pr is the fraction of production directly transferred to the root; Rs , Rr , and Re are the shoot, root, and epiphyton respiration (day
1), respectively; Ls , Lr , and Le are the sloughing or non-respiration losses for shoot, root, and epiphyton (day
1), respectively; Jrs is the carbon transport, positive from root to shoot (g C m
2 day
1); and Pe is the epiphyton production (day
1). The production rate for shoots may be expressed as Eq. (5). Light availability often plays a key role in submerged macrophyte development. The growth limiting function for light intensity, f 2 ðIÞ, was calculated using the Steele equation (Steele, 1962) in the earlier model. Because light varies continuously with time, some models integrated the light limitation function over 24 h to obtain a daily average value. Cerco and Cole (1994) and Park et al. (1995) used the daily and layer-integrated form of f 2 ðIÞ in their eutrophication models. Then, considering the average height of submerged macrophyte shoots above the bed, Ji (2008) described the light limitation as Eq. (6) in the SAVM. Ps ¼ PMs  f 1 ðNÞ  f 2 ðIÞ  f 3 ðTÞ  f 4 ðSÞ  f 5 ðDÞ

f 2 ðIÞ ¼

2:718

me  Hs

ðexpð
aÞ
expð
bÞÞ

(5)

(6)

a¼

Ii  expð
me HÞ Iso

(6a)

b¼

Ii  expð
me ðH
Hs ÞÞ Iso

(6b)

me ¼ m0 þ mTSS þ ma

(6c)

where PMs is the maximum growth rate under optimal conditions for shoots (day
1); f 1 ðNÞ, f 2 ðIÞ, f 3 ðTÞ, f 4 ðSÞ, and f 5 ðDÞ are the

126

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

me ¼

C Be þ Ke SD CChle

(7)

where SD is the Secchi disk depth (m); C is a dimensionless constant; K e is the coefficient for epiphytic chlorophyll-a (m
1 mg
1 m2); and CChle is the carbon/chlorophyll-a ratio for epiphyton (g C:mg chlorophyll-a). The water transparency was estimated by Secchi depth in the paper. me is confirmed only by the term C=SD as one scheme (without epiphyton) in our study. TSS and algae can influence water transparency. In the existing SAVM, the me was computed by the TSS and algae concentrations, in addition to the background constant (Eq. (6c)) (Jin and Ji, 2005). Thus, the me was calculated by Eq. (7) because the effect of TSS and algae on turbidity should be included in the observations of Secchi depth in M-SAVM. Then, the light attenuation coefficient varies with spatial and temporal variations of SD data, rather than primarily depending on the background constant. Because TSS is rather lower (1–4 mg l
1) during study period, SD data may be more appropriate for simulation in our study. Furthermore, the advantage of modification is that the measured data of water transparency (Secchi depth) is more readily available. Constant C in Eq. (7) has values of 1.7–1.9 in Lake Huron (Beeton, 1958), 1.33 in Chesapeake Bay (Cerco and Moore, 2001), and 1.83 in Lake Okeechobee (Ji, 2008). In the Yuqiao Reservoir, the parameter C is found to be 1.07 by measurements of light penetration evaluated from six datasets of sunny, cloudy, and rainy days in 2008 and 2009, respectively. 2.5. Coupling M-SAVM and YRWQM The Yuqiao Reservoir Water Quality Model (YRWQM) was developed under the framework of the EFDC model to simulate hydrodynamics and water quality in the Yuqiao Reservoir. YRWQM, which has been previously described, can predict acceptable water temperature (RMSE of 1.27  C) and nutrient

2.6.1. Framework of the simulations YRWQM divided the horizontal plane of the reservoir into 2329 cells (200 m  200 m) with four layers. Considering the settling velocity for state variables in the vertical coordinate, the time step was dependent on the small vertical dimension of the cells in the kinetic equations of YRWQM (see Zhang et al., 2013 for details). The time step in YRWQM was 10 s, which was in compliance with the Courant–Friendrich–Levy (CFL) criterion. Therefore, M-SAVM was coupled with YRWQM at the same time step. The initial condition was pre-proceeded in YRWQM ahead of simulation time and later put into the model based on the observation data of the first day. Both initial shoot and root carbon biomass values of the P. crispus component were assumed to be 5 g C m
2 in the bottom layer. Hydraulic (Fig. 2) and meteorological boundary conditions were all daily averaged data and water quality and loads on a semi-monthly basis (Zhang et al., 2013). As input to M-SAVM, the Secchi depth data are interpolated and extrapolated to each model cell based on the observations at Stations S1–S7. Interpolation and extrapolation were performed by the inverse distance weighted method using nearly 16 points. The SD time series data of each cell are composed of 18 sets of observed

a 120 3

In Lambert-Beer’s law, the light attenuation coefficient, me , is a measure for the reduction (absorption) of light intensity when it penetrates a medium. Then, the light attenuation is related to water transparency, and considering the effect of epiphyton, me may be expressed as Eq. (7) in M-SAVM.

2.6. Model application

Discharge (m /s)

2.4. Modified light attenuation equation in M-SAVM

concentrations (TP, RMSE of 0.01 mg l
1) (see Zhang et al., 2013 for details). These factors are essential for predicting biomass growth. For M-SAVM (or the SAVM), the submerged macrophyte is assumed to have a fixed nutrient composition. Nitrogen and phosphorus in biomass are represented in terms of carbonaceous biomass (using the parameters anc , apc ). Submerged macrophyte respiration releases nutrients back to the water column and sediment bed. M-SAVM is directly coupled with YRWQM: the growth and decay of submerged macrophyte are linked to the nutrient pool of the water quality model; the photosynthesis and respiration of submerged macrophyte link to dissolved oxygen dynamics; the settling of particulate organic matter and nutrient uptake affects nutrients in the water and sediment; and the shoot detritus is coupled with the water quality model only at the bottom layer. All state variables are updated at each model time step. The equations are detailed in Ji’s book (Ji, 2008).

2008 inflow 2008 outflow 2009 inflow 2009 outflow

100 80 60 40 20 0 0

30

60 90 Days from 1/3/2008 (2009)

120

150

b 14 Maximum Wind speed (m/s)

nutrient, light intensity, temperature, salinity, and shoot selfshading limitation coefficients, respectively (each limitation coefficient is between 0 and 1); me is the light attenuation coefficient (m
1); Hs is the average shoot height above the bed (m); H is the water depth (m); Ii is the instantaneous light intensity at the water surface (Langley day
1); ISO is the optimal light intensity on the shoot surface for submerged macrophyte growth (Langley day
1); m0 is the background light attenuation coefficient (m
1); and mTSS and ma are the light attenuation coefficients for total suspended solids (TSS) and algae chlorophyll-a (m
1), respectively. Submerged macrophyte is capable of absorbing nutrients from either the sediment or the water column. Nutrient limitation for nitrogen and phosphorus is evaluated using Madden and Kemp’s (1996) equation. The growth limiting functions for temperature and shoot self-shading are considered in our study. The effect of temperature and shoot self-shading on shoot growth is given by Ji (2008). The formulae of the respiration rate for shoot, root, epiphyton, and the carbon transport were also detailed in the literature (Ji, 2008). Nutrient enrichment also enhances epiphytic growth on submerged macrophyte leaf surfaces, which can limit the light available for submerged macrophyte photosynthesis. Epiphytic production is computed using a formula similar to Eq. (5).

12

M±SE = 4.23±0.05

10 8 6 4 2 0

0

90

180

270 360 450 Days from 1/1/2008

540

630

720

Fig. 2. Hydraulic boundary conditions in the model and maximum wind velocity time series in 2008 and 2009.

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

127

data, which are interpolated in the time domain in M-SAVM. The maximum wind speed (Fig. 2) shows that strong wind does not often occur in the reservoir and the observed SD data are appropriate for the simulation.

Similarly, water was transferred 30 days earlier from March 27 to May 19 in Sc5.

2.6.2. Calibration and validation M-SAVM is calibrated and verified with the biomass, using two datasets from March 1–July 18 in 2008 and 2009. Statistical methods such as the standard error of estimate, constrained regression analysis, absolute error, root mean squared error (RMSE), and relative root mean squared error (R-RMSE) are used for calibration and validation. The calibrated values of the major coefficients in M-SAVM are given in Table 1. Examples of model calibration and validation results of biomass at Stations S2 and S4–S7 are depicted in Fig. 3. Fig. 3 demonstrates that simulated temporal variations of biomass are consistent with the trends of measured data at each station. Table 2 summarises the results of the statistical error quantification methods for calibration and validation of biomass at Stations S2 and S4–S7. The model simulations are acceptable, based on the standard error of the estimate, absolute error, and RMSE, and the R-RMSE values are lower than 20% (Table 2).

3.1. The biomass and distribution of P. crispus influenced by epiphyton and water depth

2.6.3. Five hydraulic scenarios Because reservoir operation could not be easily changed in reality, five hydraulic scenarios (Fig. 4) were simulated by M-SAVM. The original scenario (OS) was the validation in 2009, initial water level (WL) was 19.63 m; the water was transferred from the Li River to the Yuqiao Reservoir during the period of April 26–June 18. The scenarios were classified into two types: changing water levels (Sc1–Sc3) and the beginning of the water transfer (Sc4 and Sc5). Sc1 decreased WL by 0.5 m, Sc2 and Sc3 increased WL by 0.5 m and 1.0 m, respectively. The other conditions in Sc1– Sc3 were similar to OS. For Sc4, water was transferred 15 days earlier from April 11 to June 3 with inflow adjusted accordingly.

3. Results

The SAVM for the Yuqiao Reservoir is simulated with a background light attenuation coefficient constant (0.475 m
1) to compare with the results of M-SAVM with epiphyton (Fig. 3). The results of the two models show that M-SAVM improves the accuracy in which R-RMSE decreases by 4.7% (Table 3). The verified results show the modeled biomass in the order Stations S7 > S5 > S2 > S6 > S4 (Fig. 3). The maximum of each station appears on May 12 or 13, 2009. The result in the late growth period is slightly high for simulation without epiphyton (M-SAVM lines in Fig. 3). The biomass is slightly lower (2.63% decrease, Fig. 3) in the model with epiphyton because its increased suppression of plant growth. M-SAVM with epiphyton behaves accurately 0.9% (R-RMSE) higher than that without epiphyton (Table 3). The light intensity limitation coefficient in M-SAVM with epiphyton ranges from 0.123 to 0.890, with an average of 0.381, which is 5.5% lower than that without epiphyton (0.125
0.937, 0.403  0.020) (Fig. 5). The simulations reveal a highly significant relationship (p < 0.01) between the biomass of epiphyton and P. crispus (Fig. 6). The modeled spatial distribution of P. crispus in different growth periods is presented in Fig. 7. The spatial distribution of P. crispus shows that substantial abundance is simulated in the northern and southern parts of the reservoir. A lesser amount is simulated in the western part of the reservoir. Analysing the spatial variation of P. crispus abundance and found that substantial abundance is found on both sides of the original river channel with banded distribution. The results show that substantial abundance is

Table 1 Major M-SAVM coefficients and constants of the calibrated model.a Parameter

Description

f Pr Ls ,Lr ,Le PMs

Fraction of production directly transferred to the root Non-respiration loss rate for plant shoots, roots, and epiphytes Maximum growth rate under optimal conditions for plant shoots

0.15 0.01 0.17

*, *,,KNhb

Maximum growth rate under optimal conditions for epiphytes Half-saturation constant for nitrogen uptake from water and bed

0.065 0.1

* ,,KPhb

Half-saturation constant for phosphorus uptake from water and bed

0.03

ISO *, *, *,

Maximum value for optimum light intensity for plant growth Effect of temperature below optimal low temperature on shoot production Effect of temperature above optimal high temperature on shoot production Attenuation due to shoots self-shading

*,

Maximum respiration rate for plant shoots at reference temperature

0.02

*,

Maximum respiration rate for plant roots at reference temperature

0.01

*, * ,,K TRr *, *, Ke Hs anc apc CChle

Maximum respiration rate for epiphytes at reference temperature Effect of temperature on plant shoots and roots respiration Root-to-shoot transfer rate for light dependence Half-saturation light at shoot surface Coefficient for epiphyte chlorophyll The average shoot height above the bed Nitrogen to carbon ratio for plant shoots, roots, and epiphytes Phosphorus to carbon ratio for plant shoots, roots, and epiphytes Carbon to chlorophyll ratio for epiphytes

a

Value

90 0.001 0.001 0.02

0.0005 0.001 0.35 120 0.04 1.2 0.167 0.007 0.12

Reported in literature

Unit

0.15 Ji (2008) 0.01 Ji (2008) 0.1 Cerco and Moore (2001) 0.2 Ji (2008) 0.2-0.3 Janse et al. (2010) 0.144 Van Nes et al. (2002) – 0.13 Thursby and Harlin (1982) 0.19 Cerco and Moore (2001) 0.02 Madden and Kemp (1996) 0.028 Cerco and Moore (2001) 20-100 Ji (2008) 0.001 Ji (2008) 0.001 Ji (2008) 0.045 Cerco and Moore (2001) 0.0143 Ji (2008) 0.019 Madsen et al. (1991) 0.022 Cerco and Moore (2001) 0.022 Cerco and Moore (2001) 0.01 Ji (2008) – 0.001 Ji (2008) – – – – 0.167 Cerco and Cole (1994) 0.007 Cerco et al. (2004) 0.15 Mao et al. (2009)

– day
1 day
1

The parameters with the symbol ‘*’ are in the equations for f 1 ðNÞ,f 3 ðTÞ, f 5 ðDÞ, Pe , Rs , Rr , Re , and Jrs , which are detailed elsewhere (Ji, 2008).

day
1 g N m
3 g P m
3 W m
2 day
1 
2 C 
2 C m2 g
1 day
1 day
1 day
1 
1 C day
1 W m
2 day
1 m
1 mg
1 m2 m g N: g C g P: g C g C: mg chlorophyll-a

128

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

a

b

400

300 Measured SAVM

300

M-SAVM with epiphyton M-SAVM without epiphyton

S2

S2

250 200 150

200

100

100

50

0 250

S4

0 150

S4

200 100

150 100

50

-2

Biomass (g DM m )

50 0 300

S5

250

0 250

S5

200

200

150

150

100

100

50

50 0 250

S6

200

0 200

S6

150

150

100

100 50

50 0 500

S7

0 500

400

400

300

300

200

200

100

100

S7

0

0 0

30

60

90

120

150

0

30

60

90

120

150

Days from 1/3/2009

Days from 1/3/2008

Fig. 3. Measured and modeled biomass (dry mass) at Station S2 and Stations S4–S7 for the calibration and validation.

Table 2 Statistical error quantification analysis results of calibration and validation for biomass at stations.a Parameter

Five stations

S2

S4

S5

S6

S7

i. Calibration period Number of measurements Mean of measurements Standard error of estimate Slope Absolute error RMSE R-RMSE (%)

44 241.2 32.93 1.33 30.15 41.84 17.3

9 214.7 33.20 1.11 26.79 29.97 14.0

8 143.5 21.47 0.97 17.91 21.36 14.9

11 218.9 23.84 0.59 26.86 31.56 14.4

5 196.6 3.07 1.21 15.32 16.05 8.2

11 432.2 25.12 0.48 63.89 69.45 16.1

ii. Validation period Number of measurements Mean of measurements Standard error of estimate Slope Absolute error RMSE R-RMSE (%)

46 194.9 30.77 0.94 23.23 30.64 15.7

9 168.7 30.32 1.52 26.33 29.28 17.4

10 129.0 16.18 1.04 11.93 14.78 11.5

11 185.3 24.74 1.14 22.55 28.15 15.2

4 147.7 28.94 1.44 18.51 21.80 14.8

12 343.9 43.46 1.39 36.83 43.47 12.6

a

R-RMSE: RMSE/Mean.

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

3.2. Relationship between biomass and water depth in five scenarios simulation

Water Level (m)

25 OS Sc2 Sc4

23

Sc1 Sc3 Sc5

21 19 17 15 30

0

60 90 Days from 1/3/2009

120

150

Fig. 4. Water level of the original scenario and five scenarios.

Table 3 Statistical results for biomass of M-SAVM and the SAVM. Parameter

Biomass

Number of measurements 46 Mean of measurements 194.9 i. The SAVM ii. M-SAVM With epiphyton 30.77 Standard error of estimate 33.44 Slope 0.81 0.94 Absolute error 26.88 23.23 RMSE 39.85 30.64 R-RMSE (%) 20.4 15.7

129

The biomass increases under low water levels due to increasing underwater light intensity (Sc1) and decreases when the water level is raised (Sc2–Sc5). Highly significant differences (p < 0.001) are simulated for all scenarios, except Sc4, at most stations (Fig. 8). A 0.5-m decline in water level (Sc1) results in a 21.3% increase in maximum biomass. The reduced biomass of Sc3 ranks first, along with the reduction biomass order of Sc5 > Sc2 > Sc4. Sc3 and Sc4 result in a loss of 58.4% and 17.2% P. crispus habitat, respectively. For all scenarios, the change ratio of all stations maximizes at Station S6, and the minimum appears at Station S7. The biomass of Station S7 is relatively high with maximums of 360.69–403.02 g DM m
2 in five scenarios. Based on the simulated regression (Fig. 9) in the reservoir, an empirical formula between biomass and water depth is derived in Eq. (8). Bmax ¼
64:78H þ 526:17

(8)
2

where Bmax is the maximum biomass (g DM m ). The application of the formula is suitable for H < 10 m and Bmax < 450 g DM m
2. The maximal biomass of P. crispus appears at depths 1–3 m (Fig. 9) in the reservoir.

Without epiphyton 30.95 0.91 23.53 32.34 16.6

4. Discussion 4.1. Comparison among M-SAVM, the SAVM and other existing macrophyte models

f1 (N) , f2 (I) , f3(T)

1.0

N I T I-no epiphyton I-SAVM

0.8 0.6 0.4 0.2 0.0 30

0

60 90 Days from 1/3/2009

120

150

Fig. 5. Growth limiting functions in M-SAVM for nutrients, light intensity, and temperature at Station S2 in the year 2009.

Log (P. crispus biomass (g DM m-2 ))

related to water depth. There are smaller amounts near the shore side and the deeper part of the reservoir. The modeled and measured biomass at Stations S4 and S6 are both smaller than other stations, with deeper depth of 6.26 m and 6.12 m, respectively.

3.0 y = -0.0087x + 2.296 r = -0.43

2.5 2.0 1.5 1.0

Jones and Sayer's regress ion line y = -0.0085x + 0.92

0.5 0.0

0

2

4

6

8

10

12

14

16

18

20

-2

Epiphyton biomass (mg chlorophyll-a m ) Fig. 6. Relationship between the biomass of epiphyton and P. crispus at Station S2. Black points are the measurements. Double dashed line is the part of the regression line proposed by Jones and Sayer (2003) in 17 plant-dominated shallow lakes in Norfolk (UK).

The difference in temporal biomass trends between M-SAVM and the SAVM shows a lower biomass in the decay period (June and July) in M-SAVM, which matches considerably well with observations. The biomass calculated by the SAVM in the decay period is relatively high, a drawback that has been overcome by the M-SAVM. The reason is that in the modified light attenuation equation, the light attenuation coefficient is a function of water transparency, rather than depending on the background constant. Accompanied by submerged macrophyte sloughing, the water transparency decreased by 50% during this period, and then an increased light attenuation coefficient reduced underwater light intensity and photosynthesis. The light intensity limitation coefficient of submerged macrophyte growth in M-SAVM is lower than the SAVM during this period. This relationship suggests that the result modeled by M-SAVM is more reasonable. Other macrophyte models, such as SAGA (Hootsmans, 1999), PCLake (Janse, 2005; Janse et al., 2010), and Charisma (Van Nes et al., 2002), are developed with light attenuation equations available. PCLake and Charisma performed well with a background extinction or turbidity of 0.5 m
1 in light attenuation equations. However, the modeled total biomass of Potamogeton pectinatus is higher than the observed value in the decay period in Lake Veluwemeer in the Netherlands by SAGA. That light attenuation coefficient is increasingly homogenized due to the background constant, the same as with the SAVM, which could partly explain the higher modeled biomass values. Of the many different types of parameters in these models, the maximum growth rate of vegetation is another important parameter besides the light attenuation coefficient. Comparing values demonstrates that our calibrated PMs (0.17 day
1) is appropriate for the simulation. Compared with the model results of Chara aspera biomass in Lake Veluwemeer by Charisma, the M-SAVM results are in accordance with the equilibrium rule of biomass and water depth at different levels of the light attenuation coefficient. The threshold of water depth available for vegetation growth decreases with increasing light attenuation (Van Nes et al., 2002). In their

130

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

Fig. 7. Simulated spatial distribution of P. crispus biomass for the reservoir in different periods: (a) period of seedling establishment; (b) period of exponential growth; (c) period of budding; (d) period of plant sloughing; (e) period of dying out.

presented study, the vegetation disappears at a critical water level with different light attenuation (e.g., H = 4.5 m when me = 1 m
1; H = 2.2 m when me = 3 m
1). Therefore, it is reasonable that plants grow below the water depth of 6 m with a range of me from 0.39 m
1 to 2.83 m
1 in M-SAVM in the reservoir. In addition, our results support the viewpoint of Van Nes et al. (2002) that the slightest initial seed bank (e.g., initial biomass of 5 g C m
2) is sufficient to trigger recovery under good clear conditions (me < 4 m
1) in shallow water. 4.2. Role of epiphyton for light attenuation and submerged macrophyte development

phytoplankton under conditions of submerged macrophyte dominance (Jones et al., 2002). Contrary to the theory proposed by Phillips et al. (1978), epiphyton biomass was not related to nutrient concentration but to the density of grazing invertebrates in their regions (Brönmark and Weisner 1992; Jones et al., 1998, 2002; Jones and Sayer, 2003; Beresford and Jones, 2010). In this study, the value of f 1 ðNÞ (0.97) suggests that adequate nutrients can be supplied for submerged macrophyte and epiphyton growth in the reservoir. The result indicates that the main constraint on epiphyton biomass is not nutrients. 4.3. Role of water depth for submerged macrophyte development

The simulations reveal a highly significant relationship (p < 0.01) between the biomass of epiphyton and P. crispus, which is similar to the results of measured data in Jones and Sayer’s (2003). The measured epiphyton and similar slopes of the two regression lines both suggest that the calibrated parameters (PMe , K e , CChle ) for epiphyton are appropriate for the simulation. The growth limiting function for light intensity is identified as a primary controlling factor for submerged macrophyte and epiphyton growth (Sand-Jensen, 1977). The difference in f 2 ðIÞ reveals that epiphyton plays an important role in light attenuation in the reservoir. The results indicate that epiphyton has a certain effect on the P. crispus growth in the reservoir. Indeed, epiphyton appeared to have a stronger influence on plant growth than

Reservoir differs from natural lake because the former offers a unique chance for submerged macrophyte management by changing the hydrological regimes manually. A significant negative correlation is found between the maximum biomass and water depth in the reservoir in five hydraulic scenarios. The results suggest that P. crispus grows at a water depth of 1–6 m in the reservoir. The probable reason is that increased light intensity with reduced depth enhances biomass. The reason for minimal existing P. crispus in the shallower area (shore side) is attributed to a low water depth (
C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

scenarios, The P. crispus reach maximal biomass at a depth between 1.08 m and 2.83 m at Station S7, therefore no significant changes occurred except Sc3. WLFs have an impact on light available for submerged macrophytes, which could be a method of managing excessive biomass in shallow lakes (Coops et al., 2003; Paillisson and Marion, 2011). Contrary to Sc1, biomass decreases significantly during the growth period when the water depth is increased (Sc2 and Sc3). It decreases significantly only after water diversion operation, before which the water level is unchanged (Sc4 and Sc5). This result suggests that manipulation by fluctuating water levels in a shallow reservoir or lake can significantly control the biomass of submerged macrophytes. Indeed, for two lakes in southern Sweden, it was shown that between-year trends of increasing macrophyte coverage corresponded with low water levels and decreasing coverage corresponded with high water levels (Blindow et al., 1993).

300

S2

OS Sc1 Sc2 Sc3 Sc4 Sc5

250 200 150

p1 = 0.000 p2 = 0.000 p3 = 0.000 p4 = 0.097

100

p5 = 0.000

50 0 250

S4 p1 = 0.000

200

p2 = 0.000

150

p3 = 0.000 p4 = 0.013

-2

Biomass (g DM m )

100

p5 = 0.000

50 0 250

S5

200

p1 = 0.000

150

p3 = 0.000

131

p2 = 0.000

5. Conclusions

p4 = 0.083

100

p5 = 0.000

SAVM is developed into the M-SAVM by modification of the light attenuation equation. Considering the effect of epiphyton in M-SAVM, epiphyton increase has a slightly low light intensity limitation coefficient to suppress plant growth. To a certain extent, epiphyton plays an important role in light attenuation and affects submerged macrophyte development in the reservoir. Results demonstrate a significant negative correlation between biomass and water depth in the reservoir. The biomass increases under low water levels due to increasing underwater light intensity and decreases when the water level is raised. As a feasible strategy, the hydraulic control (WLFs) can be accomplished by M-SAVM that could be useful in the management of submerged macrophytes in a shallow reservoir or a shallow lake.

50 0 250

S6

200

p1 = 0.000 p2 = 0.000

150

p3 = 0.000 p4 = 0.003

100

p5 = 0.000

50 0 500

S7

400

p1 = 0.986 p2 = 0.293

300

p3 = 0.000 p4 = 0.902

200

p5 = 0.155

Acknowledgements

100 0 0

30

60

90

120

150

Days from 1/3/2009 Fig. 8. Modeled biomass at Station S2 and Stations S4–S7 for the original scenario and five scenarios. p value for Tukey’s HSD test with 95% confidence intervals. Values in bold indicate a significant difference in biomass between each hydraulic scenario and OS.

This research was supported by the National Natural Science Foundation of China (No. 50909070), the Science Fund for Creative Research Groups of the National Natural Science Foundation of China (No. 51321065) and the Tianjin Municipal Natural Science Foundation (No. 13JCQNJC09200). Special thanks are addressed to the High Performance Computing Center of Tianjin University, China. References

Biomass (g DM m -2)

500 400

y = -64.78x + 526.17 r = -0.97

300 OS Sc1 Sc2 Sc3 Sc4 Sc5

200 100 0

regres

0

1

2

3

4

5

6

7

8

9

10

Depth (m) Fig. 9. Relationship between biomass and water depth in the reservoir. The biomass is a maximum value during the growth period at each station and at the corresponding water depth.

vegetation occurs in both shallower (shore side) and deeper (varying with light attenuation) areas (Van Nes et al., 2002). The maximal biomass of P. crispus appearance at depths (1–3 m) in the reservoir is probably due to the differences in geographical location, watershed features, and macrophyte species. For all

Beeton, A.M., 1958. Relationship between Secchi disk readings and light penetration in Lake Huron. Trans. Am. Fish. Soc. 87, 73–79. Beresford, A.L., Jones, J.I., 2010. Weedbeds and big bugs: the importance of scale in detecting the influence of nutrients and predation on macroinvertebrates in plant-dominated shallow lakes. Freshwater Bio. 55, 514–530. Berger, C.J., Wells, S.A., 2008. Modelling the effects of macrophytes on hydrodynamics. J. Environ. Eng. ASCE 134, 778–788. Blindow, I., Andersson, G., Hargeby, A., Johansson, S., 1993. Long-term patterns of alternative stable states in two shallow eutrophic lakes. Freshwater Biol. 30, 159–167. Boesch, D.F., Brinsfield, R.B., Magnien, R.E., 2001. Chesapeake Bay eutrophication: scientific understanding ecosystem restoration and challenges for agriculture. J. Environ. Qual. 30, 303–320. Brönmark, C., Weisner, S.E.B., 1992. Indirect effects of fish community structure on submerged vegetation in shallow, eutrophic lakes: an alternative mechanism. Hydrobiologia 243–244, 293–301. Cerco, C.F., Cole, T.M., 1993. Three-dimensional eutrophication model of Chesapeake Bay. J. Environ. Eng. ASCE 119, 1006–1025. Cerco, C.F., Cole, T.M., 1994. Three-dimensional eutropication model of Chesapeake Bay. Volume I: Main Report. Technical Report EL-94-4. U.S. Army Corps of Engineers. Waterways Experiment Station, Vicksburg, MS. Cerco, C.F., Linker, L., Sweeney, J., Shenk, G., Butt, A.J., 2002. Nutrient and solids controls in Virginia’s Chesapeake Bay tributaries. J. Water Res. Plann. Manage. ASCE 128, 179–189. Cerco, C.F., Moore, K., 2001. System-wide submerged aquatic vegetation model for Chesapeake Bay. Estuaries 24, 522–534.

132

C. Zhang et al. / Ecological Engineering 81 (2015) 123–132

Cerco, C.F., Noel, M.R., Tillman, D.H., 2004. A practical application of Droop nutrient kinetics (WR 1883). Water Res. 38, 4446–4454. Chen, Y.Y., Zhang, C., Gao, X.P., Wang, L.Y., 2012. Long-term variations of water quality in a reservoir in China. Water Sci. Technol. 65, 1454–1460. Cole, T.M., Buchak, E.M., 1995. CE-QUAL-W2: a two dimensional, laterally averaged hydrodynamic and water quality model, Version 2.0, user manual. Instruction Report EL-95-1. U.S. Army Corps of Engineers. Waterways Experiment Station, Vicksburg, MS. Coops, H., Hosper, H., 2002. Water-level management as a tool for the restoration of shallow lakes in the Netherlands. Lake Reservoir Manage. 18, 293–298. Coops, H., Beklioglu, M., Crisman, T.M., 2003. The role of water-level fluctuations in shallow lake ecosystems – workshop conclusions. Hydrobiologia 506–509, 23–27. Deegan, L.A., Wright, A., Ayvazian, S.G., Finn, J.T., Golden, H., Merson, R.R., Harrison, J., 2002. Nitrogen loading alters seagrass ecosystem structure and support of higher trophic levels. Aquat. Conserv. Marine Freshwater Ecosyst. 12, 193–212. DHI, 2001. MIKE 3 Estuarine and coastal hydrodynamics and oceanography. DHI Water & Environment. Danish Hydraulic Institute, Horsholm, Denmark. Dierberg, F.E., Juston, J.J., DeBusk, T.A., Pietro, K., Gu, B.H., 2005. Relationship between hydraulic efficiency and phosphorus removal in a submerged aquatic vegetation-dominated treatment wetland. Ecol. Eng. 25, 9–23. Gulati, R.D., Van Donk, E., 2002. Lakes in the Netherlands, their origin, eutrophication and restoration state-of-the-art review. Hydrobiologia 478, 73–106. Hamrick, J.M., 1992. A Three-Dimensional Environmental Fluid Dynamics Computer Code: Theoretical and Computational Aspects. The College of William and Mary, Virginia Institute of Marine Science Special Report 317, 63 pp.. Hilt, S., Gross, E.M., Hupfer, M., Morscheid, H., Mahlmann, J., Melzer, A., Poltz, J., Sandrock, S., Scharf, Eva-M., Schneider, S., Van de Weyer, K., 2006. Restoration of submerged vegetation in shallow eutrophic lakes – a guideline and state of the art in Germany. Limnologica 36, 155–171. Hootsmans, M.J.M., 1994. A growth analysis model for Potamogeton pectinatus. Geobotany 21, 250–286. Hootsmans, M.J.M., 1999. Modelling Potamogeton pectinatus: for better or for worse. Hydrobiologia 415, 7–11. Janse, J.H., 1997. A model of nutrient dynamics in shallow lakes in relation to multiple stable states. Hydrobiologia 342/343, 1–8. Janse, J.H., 2005. Model studies on the eutrophication of shallow lakes and ditches. Ph.D. Thesis. Wageningen University, the Netherlands Available on line at http://edepot.wur.nl/121663.pdf. Janse, J.H., De Senerpont Domis, L.N., Scheffer, M., Lijklema, L., Van Liere, L., Klinge, M., Mooij, W.M., 2008. Critical phosphorus loading of different types of shallow lakes and the consequences for management estimated with the ecosystem model PCLake. Limnologica 38, 203–219. Janse, J.H., Scheffer, M., Lijklema, L., Van Liere, L., Sloot, J.S., Mooij, W.M., 2010. Estimating the critical phosphorus loading of shallow lakes with the ecosystem model PCLake Sensitivity, calibration and uncertainty. Ecol. Modell. 221, 654–665. Jeppesen, E., Jensen, J.P., Søndergaard, M., Lauridsen, T., Pedersen, L.J., Jensen, L., 1997. Top-down control in freshwater lakes: the role of nutrient state, submerged macrophytes and water depth. Hydrobiologia 342–343, 151–164. Ji, Z.G., 2008. Hydrodynamics and Water Quality: Modeling Rivers, Lakes, and Estuaries. John Wiley & Sons, Inc., Hoboken, New Jersey. Jin, K.R., Ji, Z.G., 2005. Application and validation of three-dimensional model in a shallow lake. J. Waterw. Port Coastal Ocean Eng. ASCE 131, 213–225. Jin, K.R., Ji, Z.G., James, R.T., 2007. Three-dimensional water quality and SAV modeling of a large shallow lake. J. Great Lakes Res. 33, 28–45. Jin, K.R., Ji, Z.G., 2013. A long term calibration and verification of a submerged aquatic vegetation model for Lake Okeechobee. Ecol. Processes 2, 23. Jones, J.I., Moss, B., Young, J.O., 1998. The interactions between periphyton, nonmolluscan invertebrates, and fish in standing freshwaters. In: Jeppesen, E., Søndergaard, M., Christophersen, K. (Eds.), The Structuring Role of Submerged Macrophytes in Lakes. Springer-Verlag, New York, pp. 69–90. Jones, J.I., Sayer, C.D., 2003. Does the fish-invertebrate-periphyton cascade precipitate plant loss in shallow lakes? Ecology 84, 2155–2167. Jones, J.I., Young, J.O., Eaton, J.W., Moss, B., 2002. The influence of nutrient loading: dissolved inorganic carbon and higher trophic levels on the interaction between submerged plants and periphyton. J. Ecol. 90, 12–24. Knight, R.L., Gu, B.H., Clarke, R.A., Newman, J.M., 2003. Long-term phosphorus removal in Florida aquatic systems dominated by submerged aquatic vegetation. Ecol. Eng. 20, 45–63. Kosten, S., Kamarainen, A., Jeppesen, E., Van Nes, E.H., Peeters, E.T.H.M., Mazzeo, N., Sass, L., Hauxwell, J., Hansel-Welch, N., Lauridsen, T.L., Søndergaard, M., Bachmann, R.W., Lacerot, G., Scheffer, M., 2009. Climate-related differences in the dominance of submerged macrophytes in shallow lakes. Global Change Biol. 15, 2503–2517. Leira, M., Cantonati, M., 2008. Effects of water-level fluctuations on lakes: an annotated bibliography. Hydrobiologia 613, 171–184. Li, X.Y., Weller, D.E., Gallegos, C.L., Jordan, T.E., Kim, H.C., 2007. Effects of watershed and estuarine characteristics on the abundance of submerged aquatic vegetation in Chesapeake Bay subestuaries. Estuaries Coasts 30, 840–854. Ma, H.G., Townsend, H., Zhang, X.S., Sigrist, M., Christensen, V., 2010. Using a fisheries ecosystem model with a water quality model to explore trophic and

habitat impacts on a fisheries stock: a case study of the blue crab population in the Chesapeake Bay. Ecol. Modell. 221, 997–1004. Madden, C., Kemp, W., 1996. Ecosystem model of an estuarine submersed plant community: calibration and simulation of eutrophication responses. Estuaries 19, 457–474. Madsen, J.D., Hartleb, C.F., Boylen, C.W., 1991. Photosynthetic characteristics of Myriophyllum spicatum and six submersed aquatic macrophyte species native to Lake George, New York. Freshwater Biol. 26, 233–240. Mao, J.Q., Lee, J.H.W., Choi, K.W., 2009. The extended Kalman filter for forecast of algal bloom dynamics. Water Research. 43, 4214–4224. McKellar, H.N., Tufford, D.L., Alford, M.C., Saroprayogi, P., Kelley, B.J., Morris, J.T., 2007. Tidal nitrogen exchanges across a freshwater wetland succession gradient in the upper Cooper River, South Carolina. Estuaries Coasts 30, 989–1006. Melzer, A., 1999. Aquatic macrophytes as tools for lake management. Hydrobiologia 395/396, 181–190. Ministry of water resources of the People’s Republic of China. 1996. Standard for the investigation of reservoir fishery resources. Beijing, China. Moreno, M.J., 2011. Analysis of the relationship between submerged aquatic vegetation (SAV) and water trophic status of lakes clustered in Northwestern Hillsborough County, Florida. Water Air Soil Pollut. 214, 539–546. Murphy, K.J., 2002. Plant communities and plant diversity in softwater lakes of northern Europe. Aquat. Bot. 73, 287–324. Naselli-Flores, L., Barone, R., 2005. Water-level fluctuations in Mediterranean reservoirs setting a dewatering threshold as a management tool to improve water quality. Hydrobiologia 548, 85–99. Paillisson, J.M., Marion, L., 2011. Water level fluctuations for managing excessive plant biomass in shallow lakes. Ecol. Eng. 37, 241–247. Park, K., Kuo, A.Y., Shen, J., Hamrick, J.M., 1995. A Three-dimensional Hydrodynamic Europhication Model (HEM-3D): Description of Water Quality and Sediment Processes Submodels. The College of William and Mary, Virginia Institute of Marine Science Special Report 327, 113 pp.. Phillips, G.L., Eminson, D., Moss, B., 1978. A mechanism to account for macrophyte decline in progressively eutrophicated freshwaters. Aquat. Bot. 4, 103–126. Piedade, M.T.F., Junk, W.J., Long, S.P., 1991. The productivity of the C4 grass Echinochloa polystachya on the Amazon Floodplain. Ecology 72, 1456–1463. Qin, B.Q., Song, Y.Z., Gao, G., 2006. The role of periphytes in the shift between macrophyte and phytoplankton dominated systems in a shallow, eutrophic lake (Lake Taihu, China). Sci. China Ser. C: Life Scie. 49, 597–602. Sand-Jensen, K., 1977. Effect of epiphytes on eelgrass photosynthesis. Aquat. Bot. 3, 55–63. Sayer, C.D., Burgess, A., Kari, K., Davidson, T.A., Peglar, S., Yang, H.D., Rose, N., 2010. Long-term dynamics of submerged macrophytes and algae in a small and shallow: eutrophic lake: implications for the stability of macrophytedominance. Freshwater Biol. 55, 565–583. Scheffer, M., Bakema, A.H., Wolterboer, F.G., 1993a. Megaplant: a simulation model of the dynamics of submerged plants. Aquat. Bot. 45, 341–356. Scheffer, M., Hosper, S.H., Meijer, M.L., Moss, B., Jeppesen, E., 1993b. Alternative equilibria in shallow lakes. Trends Ecol. Evol. 8, 275–279. Silva, T.S.F., Costa, M.P.F., Melack, J.M., 2010. Assessment of two biomass estimation methods for aquatic vegetation growing on the Amazon Floodplain. Aquat. Bot. 92, 161–167. Spoljar, M., Drazina, T., Sargac, J., Borojevic, K.K., Zutinic, P., 2012. Submerged macrophytes as a habitat for zooplankton development in two reservoirs of a flow-through system (Papuk Nature Park, Croatia). Ann. DE Limnologie Int. J. Limnol. 48, 161–175. Steele, J.H., 1962. Environmental control of photosynthesis in the sea. Limnol. Oceanogr. 7, 137–150. Ter Heerdt, G.N.J., Drost, H.J., 1994. Potential for the development of marsh vegetation from the seed bank after a drawdown. Biol. Conserv. 67, 1–11. Thursby, G.B., Harlin, M.M., 1982. Leaf-root interation in the uptake of ammonia by Zostera marina. Marine Biol. 72, 109–112. Van Nes, E.H., Scheffer, M., Van Den Berg, M.S., Coops, H., 2002. Dominance of charophytes in eutrophic shallow lakes – when should we expect it to be an alternative stable state? Aquat. Bot. 72, 275–296. Van Nes, E.H., Scheffer, M., Van Den Berg, M.S., Coops, H., 2003. Charisma: a spatial explicit simulation model of submerged macrophytes. Ecol. Modell. 159, 103–116. Van Nes, E.H., Van Den Berg, M.S., Clayton, J.S., Coops, H., Scheffer, M., Van Ierland, E., 1999. A simple model for evaluating the costs and benefits of aquatic macrophytes. Hydrobiologia 415, 335–339. Wang, C., Zhu, P., Wang, P.F., Zhang, W.M., 2006. Effects of aquatic vegetation on flow in the Nansi Lake and its flow velocity modeling. J. Hydrodyn. Ser. B 18, 640–648. Wetzel, R., Neckles, H., 1986. A model of Zostera marina L. Photosynthesis and growth: simulated effects of selected physical-chemical variables and biological interactions. Aquat. Bot. 26, 307–323. Wool, T.A., Ambrose, R.B., Martin, J.L., Comer, E.A., 2002. Water Quality Analysis Simulation Program (WASP), Version 6.0, User Manual. U.S. EPA. Atlanta, GA. Zhang, C., Chen, X.J., Wang, L.Y., Gao, X.P., 2011. Excessive growth of Potamogeton crispus impact on water quality and cause analysis in Yuqiao Reservoir. J. Tianjin Univ. 44, 1–6. Zhang, C., Gao, X.P., Wang, L.Y., Chen, Y.Y., 2013. Analysis of agricultural pollution by flood flow impact on water quality in a reservoir using a three-dimensional water quality modeling. J. Hydroinf. 15, 1061–1072.