Modelling ecosystem structure and trophic ...

Ecological Modelling 291 (2014) 82–95

Contents lists available at ScienceDirect

Ecological Modelling journal homepage: www.elsevier.com/locate/ecolmodel

Modelling ecosystem structure and trophic interactions in a typical cyanobacterial bloom-dominated shallow Lake Dianchi, China Kun Shan a,b , Lin Li a,∗ , Xiaoxiao Wang c , Yanlong Wu a,b , Lili Hu a,b , Gongliang Yu a , Lirong Song a,∗∗ a

State Key Laboratory of Freshwater Ecology and Biotechnology, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430072, China University of Chinese Academy of Sciences, Beijing 100049, China c Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences, Chongqing 401122, China b

a r t i c l e

i n f o

Article history: Received 28 April 2014 Received in revised form 18 July 2014 Accepted 22 July 2014 Keywords: Lake Dianchi Ecopath Food web Ecosystem property Cyanobacterial bloom Exotic fish

a b s t r a c t Lake Dianchi is the largest shallow lake in Yunnan-Guizhou plateau and the sixth largest one in China. The lake has been experiencing cyanobacterial blooms in the last two decades. Although a few studies have investigated the tempo-spatial dynamics of cyanobacterial blooms and their underlying mechanisms, knowledge regarding the food web structure and trophic interactions in bloom-dominated ecosystems is scarce. In the present study, an Ecopath model was developed to assess the entire lake ecosystem on the basis of historical and survey data obtained between 2009 and 2010 at Lake Dianchi. The results showed that the aggregation of flows sensu Lindeman refers to six trophic levels (TLs), and most biomasses and trophic flows were primarily concentrated at the first three levels. About 77.5% of the trophic flows from TLI to TLII originated from detritus, whereas high proportions of under-utilised zooplankton biomass returned to the detritus because of low transfer efficiencies (2.9%) in TLII. The microbial loop was considered to be involved in linking the transfer between detritus and TLII. In addition, low values of connectance index and average mutual information implied that the food web tended to be lost in information diversity and had a less complicated structure. High cycling flows concentrated in the microbial loop reflected that the ecosystem enhanced recycling to forms positive feedback by which ecosystem locked the nutrients and promoted the inflation of biomass in plankton communities. Thus, Dianchi Lake was clearly thought to be a bottom-up control ecosystem. These characteristics of the food web partly explained why cyanobacterial blooms were exceptionally heavy and durable in this lake. Finally, the implications of artificially stocking filter-feeding fish (bighead and silver fish) and exotic zooplantivorous icefish on the ecosystem structure and function are discussed herein. © 2014 Published by Elsevier B.V.

1. Introduction The structure, function, and stability of aquatic ecosystems are supposed to be susceptible to a wide array of human activities, including anthropogenic nutrient enrichment, over-exploited fishery activity, and alien species introduction (Wilcove et al., 1998; Pauly et al., 1998; Capriulo et al., 2002; Heisler et al., 2008). To date, early signs of climate-related changes in lake ecosystems have been

Abbreviations: EE, ecotrophic efficiency; DOC, dissolved organic carbon; POC, particulate organic carbon; AODC, acridine orange direct count; GPP, gross primary production; TL, trophic level; OI, ominivory indices; CI, connectance index; TEs, transfer efficiencies; TST, total system throughput; FCI, Finn’s cycling index; AMI, average mutual information; MTI, mixed trophic impact. ∗ Corresponding author. ∗∗ Corresponding author. Tel.: +86 027 68780806; fax: +86 027 68780871. E-mail addresses: [email protected] (L. Li), [email protected] (L. Song). http://dx.doi.org/10.1016/j.ecolmodel.2014.07.015 0304-3800/© 2014 Published by Elsevier B.V.

reported (Straile, 2002; Livingstone, 2003); ecologists have begun to pay more attention to the effects of climate variability, especially temperature increase (Carpenter et al., 1992; Garten and Adrian, 2002; Beardall and Raven, 2004; Wrona et al., 2006). The propagation of cyanobacteria has been proposed to have steadily intensified with the increase of water temperature in lakes, rivers, and reservoirs worldwide during the last decades (Carey et al., 2012; Paerl and Paul, 2012). Microcystis, the genus with the most ubiquitous and harmful blooms of cyanobacteria (CyanoHABs), blooms have widely threatened the safety and health of aquatic ecosystems (Paerl and Huisman, 2008). In the period of blooms, scums of Microcystis species accumulate on the surface of water and regulate buoyancy for the optimum utilisation of nutrients and light resources (Xiao et al., 2012); this is responsible for water quality deterioration and ecological degradation by severe changes in ecosystem properties (Carpenter et al., 1998; Paerl and Huisman, 2009). Negative impacts

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

of bloom formation and collapse are well known on aquatic ecosystems, including high turbidity and shading the light needed by aquatic plants (Berger, 1989), changing physico-chemical factors (elevated pH and reduced CO2 ), and producing toxins called microcystins (MCs) that affect the habitats of other biological communities (Kann and Smith, 1999; Hessen et al., 2005). Therefore, some studies have focused on different levels of aquatic organisms along with life cycle of Microcystis (Ke et al., 2008) and even the involved dynamics of the ‘microbial loop’ (Sommaruga, 1995) to postulate the possible relationship between cyanobacteria and community structure of bacteria and flagellates (Xing et al., 2007; Wilhelm et al., 2011). However, predicting the overall effects of blooms on ecosystem seems difficult. The community changes of different organisms can indirectly cause changes in the pathways and magnitude of material flows in food webs. These flows have been considered to reveal important features and functions of natural food webs (Berlow et al., 2004; Van Oevelen et al., 2010). In China, typical cases for Microcystis blooms are mainly found in three large shallow lakes: Lakes Taihu, Chaohu, and Dianchi; at these sites, the government has invested billions of funds over the last decade in order to prevent the occurrence of blooms. At Lake Dianchi, despite the implementation of various measures, including cutting down the point and non-point nutrients, re-establishing aquatic macrophytes in lakeside zones, and introducing artificially cultured carp (bighead and silver carp), successfully decreased the nutrient concentration to some extent (Lu et al., 2012), the Microcystis blooms still remained a severe problem. These blooms could persist for up to 10 months and cover the majority of the lake’s surface (Wan et al., 2008). Reasons for the persistence of cyanobacterial blooms and mechanisms of nutrient cycling at the ecosystem level in shallow eutrophic lakes have fascinated many ecologists (McCarthy et al., 2007; Paerl et al., 2011). Recent research has revealed the importance of internal nutrient dynamics (McDonald et al., 2010). Therefore, quantitative trophic interaction analyses at the ecosystem level are crucial for lakes with cyanobacterial blooms to better understand the effect on nutrient cycling from the food web structure. Icefish (Neosalanx taihuensis Chen) are small transparent fish that are distributed widely in both coastal and inland waters of China. It was first introduced from Yangtze River Basin to Lake Dianchi in 1979 and has rapidly dominated and become the primary economic fishery species. This species was introduced at a large scale in some Yunnan Plateau Lakes (Guo et al., 2009). Nevertheless, recently, with the widespread degeneration of lake ecosystems in Yunnan Plateau, many studies have focused on the influence of eutrophication and icefish introduction on the ecosystem structure (Liu et al., 2009; Li et al., 2011). Some ecologists have suggested that the introduction of icefish from the Yangtze River to lakes in Yunnan-Guizhou Plateau often has detrimental impacts on the ecological structure (Zhang et al., 2005). The hypothesis was based on the trophic cascading interactions that increase the zooplantivorous biomass, resulting in decreased herbivorous zooplankton biomass and increased phytoplankton biomass (Drenner and Hambright, 2002). Thus, an ecosystem-based approach is essential for the Yuannan Plateau Lakes to maintain the sustainable exploitation of exotic fisheries and health of ecosystems (Garcia, 2003; Villanueva et al., 2008). To better understand the above-mentioned issues and to assess the trophic structure and interactions in Lake Dianchi, we used the ecological network analysis (ENA) method. This analysis has been successfully used in various lakes worldwide (Moreau et al., 2001; Fetahi and Mengistou, 2007; Thapanand et al., 2007), but it is an emerging field for lake research and management in China. The most commonly used software packages for ENA are NETWRK4 (Ulanowicz, 1987) and Ecopath (Christensen and Pauly, 1992). The differences in their output are so small that modellers would

83

Fig. 1. Location of Lake Dianchi with the 24 sampling sites.

interpret the results and obtain the same qualitative conclusions (Heymans and Baird, 2000). Ecopath software was used in this study since it is user-friendly and has a standardised interface. The model definitely required large amounts of input data that directly obtained in Lake Dianchi in order to ascertain the quality of model. Our objectives were to (1) quantify the food web structure and trophic interactions in a cyanobacterial bloom-dominated ecosystem, (2) reveal ecosystem properties and development status, and (3) evaluate the ecological consequences of cyanobacterial blooms and exotic fish in Lake Dianchi. 2. Materials and methods 2.1. The study site Lake Dianchi (24◦ 29 –25◦ 28 N, 102◦ 29 –103◦ 01 E) is located at the south-west region of Kunming City, Yunnan Province, China (Fig. 1). It is a typical shallow plateau lake at an altitude of 1886.5 m, with an area of approximately 300 km2 ; mean depth, 4.7 m; and maximum depth, 11 m (Hou et al., 2004), as well as covering a total basin area of 2920 km2 . Lake Dianchi is regarded as the sixth largest freshwater lake in China and the largest lake in Yunnan Plateau, which has a distinctive monsoon climate with an annual mean temperature of 15 ◦ C; an annual mean precipitation of about 1000 mm; and annual evaporation of 1870–2120 mm (Gong et al., 2009). Lake Dianchi also provides various natural resources for local inhabitants, in terms of fishery, reserved drinking water, and tourism industry. Before the 1960s, the nutrient concentrations in Lake Dianchi were low; submerged macrophyte communities covered more than 80% area of the lake, and indigenous fish dominated the fish community. However, between the 1970s and 1980s, the amounts

84

K. Shan et al. / Ecological Modelling 291 (2014) 82–95 Table 1 Functional groups of Lake Dianchi in the Ecopath ecosystem model. Group name

Taxa

Redfin culter Common carp Bastard carp Bighead carp Silver carp Ice fish Herbivorous fishes Small fishes

Culterichthys erythropterus Cyprinus carpio Carassius auratus Aristichthys nobilis Hypophthalmichthys molitrix Neosalanx taihuensis Chen Ctenopharyngodon idellus Toxabramis swinhonis, Hyporhamphus intermedius, Rhinogobius giurinus, Pseudorasbora parva Palaemon modestu, Macrobrachium nipponense Radix swinhoe, Margarya sp. Chironomus plumosus Limnodrilus hoffmeisteri, Tubifex tubifex, Branchiura sowerbyi Microcyclops varicans, Thermocyclops taihokuensis, Mecrocyclops leuckarti, Cyclopoid nauplii Bosmina longirostris, Ceriodaphnia cornuta, Chydorus sphaericus Brachionus forficula, Keratella sp. Vorticella sp., Halteria grandinella, Strombidium sp. Potamogeton pectinatus, Potamogeton mlaianus, Myriophyllum spicatum Cyanobacteria (Microcystis sp.), Chlorophyta, Bacillariophyta Bacteria (Brevundimonas sp., Aeromonas sp., Pseudomonas sp., Sphingomonas sp.), DOC, POC

Shrimps Gastropods Chironomid larvae Oligochaetes Fig. 2. Temporal trends in fish yield and chlorophyll a in Lake Dianchi. (Historical fish data were cited from published literatures by Zhuang et al. (1996) and Peng (2002); recent landing data, from Fishery Department in Yunnan Government; Chlorophyll a data from studies by Liu (2001), Wan et al. (2008), and Li et al. (2009); other data were directly measured by Dianchi Lake Field Station of Institute of Hydrobiology, CAS.)

of macrophytes have gradually decreased along with a considerable increase in nutrient concentrations and sharp decrease in transparency from 2 to 0.6 m; indigenous fish species have declined from 26 to 13 (Qu and Li, 1983; Chen et al., 2001). The structure and biomass of phytoplankton also underwent a significant change during this period, and the biomass of cyanobacteria increased gradually and reached a peak value by the mid of the late 1990s. The ecosystem seems to have thus undergone a regime shift (Wang, 2010). Species richness of macroinvertebrate communities, as an important indicator, had significantly declined from 57 in the 1980s to merely 22 in the 1990s (Wang et al., 2011). On the other hand, when icefish was introduced, the fish yield reached the highest level throughout the 1980s. Unfortunately, the high yield was not maintained for a considerably long time. However, there was a remarkable collapse of icefish yield at the beginning of the 1990s in Lake Dianchi. At the same time, exotic piscivorous fish – redfin culter (Cultrichthys erythropterus) – began to dominate this lake (Fig. 2). 2.2. Modelling theoretical approach A static mass-balance model of Lake Dianchi was constructed using the latest version of Ecopath software (version 6.2; October 2012, freely available at http://www.ecopath.org; Chrisenten and Waters, 2004). The Ecopath model includes a set of linear equations for expressing mass-balance. Each compartment in the model is defined by two basic equations: one for the production term, and the other for the energy balance for each group. The following is the first basic equation for Ecopath: Bi

P  B

i

EEi −

n P  

Bj

j=1

B

j

DCij − Yi − Ei − BAi = 0,

(1)

where Bi and Bj are biomass (of group i and predator j); (P/B)i is the production/biomass ratio of group i; EEi is the ecotrophic efficiency of group i; (Q/B)j is the consumption/biomass ratio of predator j; DCji is the fraction of prey i in the diet of predator j; Yi is the total fishery catch rate of i, Ei is the net migration rate (emigration–immigration); and BAi is the biomass accumulation rate for i. When equalising the energy balance of a group, other flows should be considered. The second equation is as follows: Consumption = Production + Respiration + Unassimilated food. (2)

Copepods Cladocerans Rotifers Protozoa Macrophytes Phytoplankton Detritus

For each group, the model requires data and parameters consisting of B (biomass), P/B and Q/B ratio, and diet composition (DC), which can be obtained directly from both field surveys and information available in the literature. In this study, the basic input parameters were mostly obtained directly from studies performed between 2009 and 2010. Further, for balancing the model, some basic ecological and thermodynamic rules were required as follows: • The value for ecotrophic efficiency (EE) is supposed to never exceed 1.0. High value (close to 1) of EE indicates that a group is highly exploited and preyed. On the other hand, a value near 0.0 indicates a group has no predators and is not hunted. • P/Q corresponds to what is called the gross food conversion efficiency (GE); it normally ranges from 0.1 to 0.3, perhaps lower for top predators and higher for very small organisms (Ecopath User Guide, http://www.ecopath.org). • Respiration/biomass (R/B) is expected to be around 1–10·year−1 for fish and might have a greater value for species with a higher turnover. The unassimilated values need to be adjusted to obtain a suitable range of R/B (Darwall et al., 2010). 2.3. Model construction and input parameters 2.3.1. Functional groups In order to construct the mass-balance model of Lake Dianchi, we considered a total of 19 groups in this study, on the basis of diet, abundance, and information availability in the ecosystem (Christensen et al., 2000). Given the possible importance of microbial food web in Dianchi, zooplankton and zoobenthos were subdivided into four and three functional groups, respectively. Importantly, bacteria were included in the detritus group, in order to facilitate comparison with the results from other Ecopath models in literatures. Table 1 shows a list of compartments selected to represent the food web function groups in Lake Dianchi. 2.3.2. Detritus In this study, detritus was defined as non-living organic matter derived from the death of living organisms, which comprised

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

85

Table 2 Basic input values and estimated parameters (in bold) for the Ecopath model of Lake Dianchi. Group name

TL

HA (fraction)

BHA (t/km2 )

P/B

Q/B

GE

EE

OI

Redfin culter Common carp Bastard carp Bighead carp Silver carp Ice fish Herbivorous fishes Small fishes Shrimps Gastropods Chironomid larvae Oligochaetes Copepods Cladocerans Rotifers Protozoa Macrophytes Phytoplankton Detritus

3.547 2.923 2.262 2.399 2.119 3.120 2.000 2.791 3.060 2.000 2.000 2.000 2.043 2.000 2.160 2.000 1.000 1.000 1.000

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.021 1 1

7.570 0.108 1.345 0.256 0.322 3.211 0.031 1.560 29.850 4.182 1.078 7.285 10.210 36.250 9.466 2.317 1600.000 273.500 92.030

0.730 1.428 1.130 1.358 1.452 2.373 3.250 2.140 1.830 1.326 7.590 12.740 40.000 57.000 117.000 312.000 1.100 95.270 –

5.340 13.120 12.300 6.631 7.402 27.200 22.130 29.510 24.400 10.608 397.500 637.000 800.000 1140.000 390.000 1040.000 – – –

0.137 0.109 0.092 0.205 0.196 0.087 0.147 0.073 0.075 0.125 0.020 0.020 0.050 0.050 0.300 0.300 – – –

0.630 0.901 0.688 0.768 0.744 0.914 0.963 0.983 0.704 0.162 0.974 0.061 0.977 0.172 0.311 0.931 0.249 0.515 0.817

0.272 0.246 0.197 0.252 0.111 0.089 0.000 0.177 0.004 0.000 0.000 0.000 0.045 0.000 0.134 0.000 0.000 0.000 0.196

particles and dissolved organic compounds together with bacteria. The total biomass of dissolved organic carbon (DOC) and particulate organic carbon (POC) in this model was calculated using the empirical formula that was integrated into Ecoempire routine in the previous version of Ecopath (Heymans et al., 2004). log D = 0.954 log PP + 0.863 log E − 2.41,

(3) (g C m−2 );

PP reprewhere D is the total biomass of DOC and POC sents primary production (g C m−2 year−1 ); and E is the euphotic depth in metre (m). On the basis of the measurements of primary production value (4.56 g (O2 ) m−2 d−1 ) and euphotic layer depth of 0.61 m in Lake Dianchi, we calculated the sum of DOC and POC to be 9.47 g C m−2 . Assuming that organic carbon accounts for 45% of organic matter (Jørgensen et al., 1991), the unit of value was converted into wet weight (21.04 t km−2 ). Considering the hypertrophic state, Bird and Kalff’s model (Bird and Kalff, 1984) was found to be suitable for calculating the bacterial biomass; this model considers the relationship between chlorophyll a (Chla) and the abundance of bacteria (log BNUM = 5.867 + 0.776 log Chla.). With the average value of Chla (120 ␮g/L) in Lake Dianchi, the acridine orange direct count (AODC) was calculated to be 3.02 × 107 cells/mL. With the volume of bacteria estimated to be 0.5 ␮m3 (Cho and Azam, 1990), and the biomass was calculated to be 71 t km−2 . Summing up the bacterial and organic carbon values, detritus was estimated in the model to be 92.03 t km−2 . 2.3.3. Phytoplankton and primary production The biomass of phytoplankton was measured from May 2009 to April 2010 on a monthly basis at the 24 sampling stations (Fig. 1). The abundance of phytoplankton was determined by counting individuals by using an Olympus microscope (CX41), and cells were sized to derive volumes. Biomass (wet weight) was calculated assuming a wet weight density of 1 g/cm3 . The total biomass of phytoplankton varied between 36.7 and 80.9 mg/L in Lake Dianchi, with an average value of 58.2 mg/L. For the mean depth of 4.7 m, the input value in the model was 273.5 t km−2 . The gross primary production (GPP) was determined by measuring the oxygen production and consumption in water. Classic light (transparent) and dark (opaque) bottle method was used from September 2009 to August 2010 on a monthly basis by Sun et al. (2011) and Chen (unpublished data), with range values of 0.38–9.37 g (O2 ) m−2 d−1 in the water column and average value of 4.56 g (O2 ) m−2 d−1 . The value of GPP in wet weight was estimated to be 86.56 t km−2 d−1 in this model. The production/biomass (P/B) ratio for phytoplankton was calculated to be 115.52 year−1 .

2.3.4. Macrophytes There are only three communities of submerged macrophytes in Lake Dianchi (listed in Table 1). The biomass estimate of approximately 2 kg m−2 with only 2% cover in the entire lake was obtained from the sampling study by the Ecology Institute of Lake Dianchi (Pan, unpublished), during the period 2009–2010. Subsequently, the P/B value of 1.00 year−1 was chosen for the macrophytes (Han et al., 2011). 2.3.5. Zooplankton and secondary production Four major functional groups were considered in the Dianchi model: protozoa, rotifers, cladocera, and copepods. Zooplankton was sampled monthly from 2009 to 2010 at the 24 sampling stations. Zooplankton was identified and counted using a microscope. The biomass was calculated using the length-weight relationship that was used according to Huang (1999), and the mean value for each group is listed in Table 3. The zooplankton of Lake Dianchi was dominated by protozoa, rotifers, small crustaceans (mainly the Cladocera Chydorus sphaericus, Ceriodaphnia cornuta, Bosmina longirostris), cyclopoid copepods (Microcyclops varicans, Thermocyclops taihokuensis, Mecrocyclops leuckarti), and naupliar stages of copepods. An approach proposed by Kuns and Sprules (2000) (considering the mean temperature was more than 10 ◦ C and log(P/B) = −1.36–0.23 × log(w) was used to estimate the P/B ratio to be 57 for Cladocerans and 48 for Copepods, whereas the ratio was 312.2 for protozoa and 117.35 for rotifers according to the study by Liu (1999). The P/Q ratios of 0.05 and 0.3 were adopted for microzooplankton and macrozooplankton, respectively (Park, 1974; Li et al., 2009a,b). 2.3.6. Zoobenthos and benthic productivity Zoobenthos community was composed of oligochaetes, chironomids, and gastropods. The sampling was performed and calculated monthly from 2009 to 2010 at the 24 sampling stations. Oligochaetes and chironomids were sampled using Peterson’s grab sampler and filtered through a 450-␮ mesh net (Zhang and He, 1991). The gastropods were collected using a triangular haul-net with a 30-cm edge. The average biomass value for each group is listed in Table 3. Secondary production of common species in both oligochaetes and chironomids was measured using body-length frequency method consistent with the 24-site sampling in Lake Dianchi (Xie, unpublished data). Therefore, the value of P/B ratio was calculated to be 12.74 for oligochaetes, whereas it was 7.59 for chironomids.

86

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

Table 3 Diet composition of function groups in Lake Dianchi ecosystem. Group prey\predator

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Redfin culter Common carp Bastard carp Bighead carp Silver carp Ice fish Herbivorous fishes Small fishes Shrimps Gastropods Chironomid larvae Oligochaetes Copepods Cladocerans Rotifers Protozoa Macrophytes Phytoplankton Detritus Sum

2

0.030 0.493 0.056 0.026 0.394 0.001

1.000

3

0.080 0.600 0.058 0.099

4

5

6

7

8

9

10

11

12

13

14

15

16

0.091

0.001

0.059 0.100 0.050 0.039 0.010

0.009

0.515

0.153 1.000

0.227 1.000

0.104 0.172 0.043 0.069

0.009 0.011 0.027 0.068

0.001 0.093 0.093 0.252 0.337 0.002 0.002

0.547 0.362

0.463 0.289 0.248

0.010 0.020 0.010

0.160

1.000 0.164 0.448 1.000

0.236 0.649 1.000

1.000

1.000

0.129 0.091 1.000

1.000

1.000 1.000

1.000 1.000

1.000 1.000

0.450 0.510 1.000

0.200 0.800 1.000

0.330 0.510 1.000

0.110 0.890 1.000

The P/B ratio adopted in the model for gastropods was 1.326, which was obtained from the Lake Taihu model (Li et al., 2010). The P/Q ratios adopted in this model were 0.125 for gastropods and 0.02 for both oligochaetes and chironomids (Yan and Liang, 2003).

variable also expressing food type (1 for detritivores, and 0 for herbivores and carnivores).

2.3.7. Shrimps and fish The landing of fishery was obtained from Fisheries Management Department in Yunnan Province, China. A fishery survey was conducted quarterly (i.e. in February, April, July, and October) to identify fishery composition and measure the distribution of individual length and weight. For shrimps, the P/B and P/Q ratios were assumed to be 1.83 and 0.075 from the study by Halfon et al. (1996). The habitat of shrimps was assumed to cover the entire lake area. The biomass was estimated to be 29.850 t km−2 . The biomass of each fish group was estimated using the following two equations: B = Y/F and F = Z − M, where Y is the yield in t km−2 year−1 , and F, M, and Z are fishing mortality, natural mortality, and total mortality, respectively. The value of Z was considered to be equivalent to the P/B ratio under the steady condition of the ecosystem (Allen, 1971; Pauly et al., 2000). For calculation of Z, Beverton and Holt method (Beverton and Holt, 1957) was used along with FiSAT software (using ELEFAN, on the basis of lengthfrequency distribution). Natural mortality was estimated using Pauly’s empirical equation (Pauly, 1980):

The majority of diet composition for zooplankton functions was directly estimated from literatures (Dussart and Defaye, 2001; Benzie, 2005; Fetahi et al., 2011). While considering the Microcystisdominated status, the diet of Cladocerans was modified to be more detritivorous according to the stable-isotope research in the eutrophic Lake Taihu (de Kluijver et al., 2012). The diet compositions of Zoobenthos were exclusively simplified as detritivorous. Fish and shrimp diet compositions were assembled either directly from this study or from previous studies conducted on the basis of literatures on Lake Dianchi or other similar lakes (Chen et al., 1994; Liu and Zhu, 1994; Wang et al., 1995; Song, 2004; Ye, 2007; Zhou et al., 2011).

log M = −0.0066 − 0.279 log L∞ + 0.6543 log K + 0.4634 log T, (4) where L∞ (asymptotic length, cm) and K are the parameters of the Von Bertalanffy growth function, and T represents the mean annual water temperature. The Q/B ratio was calculated using the multiple regression formula of Palomares and Pauly (1998): log

Q  B

3. Results In order to overcome the uncertainty of input data, previous Ecopath software provided an ‘Ecoranger’ routine, in which it generated a theoretical frequency distribution for each basic parameter by using Monte Carlo simulations. The software could be set to run 10,000 models in order to determine the best-fitting one. The latest version of Ecopath, however, encourages users to adjust the balance of models by using ecological knowledge rather than entirely relying on computer algorithms. In this study, when all the basic inputs were integrated into the model, the diet compositions of some compartments were slightly modified in order to achieve the value of EE to be less than 1 (the modified diet matrix is shown in Table 3). Some estimated parameters computed by the model are listed in Table 2, and the trophic flow diagram is summarised in Fig. 3. 3.1. Structure analysis

= 7.964 − 0.204 log W∞ − 1.965T 

+ 0.083A + 0.532h + 0.398,

2.4. Diet composition

(5)

where W is the asymptotic weight (g), T is an expression for the mean annual temperature of the water body expressed using T = 1000/(T + 273.15). A is the aspect ratio (A = h2 (given height)/s (surface area)); h is a dummy variable expressing food type (1 for herbivores, and 0 for detritivores and carnivores), d is a dummy

3.1.1. Trophic structure and distribution of biomass The aggregation of flows referred to six trophic levels (TLs; Table 4). Trophic flows primarily occurred in the first three TLs. The producer level (TLI) mainly consisted of the detritus and phytoplankton. TLII mainly consisted of the groups of zooplankton and zoobenthos, and filter-feeding fish (silver carp and bighead carp), herbivorous fish, and some omnivore fish were partly involved

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

87

Fig. 3. Trophic flow diagram of Lake Dianchi. The direction of Y-axis is consistent with the functional TL of this box. The area of each box is proportional to the log of biomass. B is the biomass; P and Q stand for production and consumption, respectively. Flows are expressed in t km−2 y−1 . The arrows reflect direction of flows from one group to others. Solid dots at the intersections mean the connection of the two flows.

in this trophic level. Zooplanktivorous groups in terms of icefish, redfin-culter, and shrimps, as well as two zooplankton groups (rotifers and copepods), formed the first-order carnivore level (TLIII). Flows in TLIV primarily involved copepods, shrimps, and redfin culter. The latter also occupied the top trophic level in Lake Dianchi. The biomasses were also concentrated in the lower trophic levels. A relatively higher biomass in primary producers

(399.10 t km−2 ) and secondary consumers (70.70 t km−2 ) allowed the sustenance of a relatively huge proportion of predators, especially for planktivore and filter-feeding organisms. Unlike other wasp-waist ecosystems that are occupied and dominated by small plankton-feeding pelagic species (Rice, 1995), the biomass of shrimps at TLIII in Lake Dianchi was also huge (29.85 t km−2 ). Shrimps occupied more than 67% of the total fishery; however, the principal predators were

Table 4 Trophic flow matrix of Dianchi ecosystem according to groups and trophic levels. Trophic levels

Redfin culter Common carp Bastard carp Bighead carp Silver carp Ice fish Herbivorous fishes Small fishes Shrimps Gastropods Chironomid larvae Oligochaetes Copepods Cladocerans Rotifers Protozoa Macrophytes Phytoplankton Detritus Total Flow to detritus

Total

I

II

III

IV

V

VI

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 36.960 26,056.000 56,753.000 82,845.960 12,654.000

0.000 0.230 12.280 1.039 2.109 0.000 0.686 10.130 0.000 44.360 409.100 4641.000 7838.000 41,330.000 3101.000 2410.000 0.000 0.000 0.000 59,799.930 42,816.000

19.500 1.074 4.209 0.640 0.263 77.470 0.000 35.430 686.000 0.000 0.000 0.000 300.500 0.000 590.700 0.000 0.000 0.000 0.000 1715.786 1227.000

19.750 0.107 0.057 0.018 0.011 9.243 0.000 0.442 41.310 0.000 0.000 0.000 26.130 0.000 0.000 0.000 0.000 0.000 0.000 97.068 53.870

1.145 0.006 0.002 0.001 0.000 0.604 0.000 0.037 1.079 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 2.874 1.203

0.030 0.000 0.000 0.000 0.000 0.012 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.042 0.010

Note. Flow to detritus represents the proportion of the flows directly into detritus instead of consumption by predators.

40.425 1.417 16.548 1.698 2.383 87.329 0.686 46.039 728.389 44.360 409.100 4641.000 8164.630 41,330.000 3691.700 2410.000 36.960 26,056.000 56,753.000 144,461.700 56,752.810

Flow to detritus

10.128 0.511 6.265 0.590 0.834 18.119 0.285 9.265 526.024 22.394 386.238 4447.742 5316.691 28,574.540 3162.792 1616.498 27.741 12,626.150 0.000 56,752.810 –

88

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

Table 5 Distribution of flows (combined from both primary producers and detritus) and biomasses through the aggregated trophic levels. Trophic level (TL)

Catch (t/km2 )

Biomass (t/km2 )

Throughput

VI V IV III II I

0.004 0.165 3.070 19.80 1.821 0.000

0.006 0.283 5.796 38.26 70.70 399.1

0.042 2.875 97.06 1716 59,799 82,846

Consumption by predators 0.000 0.042 2.875 97.06 1716 59,799

Export 0.004 0.165 3.070 19.80 1.821 10,393

Transfer efficiency (%) 8.4 7.2 6.1 6.8 2.9

Transfer efficiencies: (1) from primary producers, 5.1%; (2) from detritus, 4.9%; total, 4.9%.

less abundant as reflected by the low biomass in the upper TLs. 3.1.2. Ominivory and connectance index The ominivory indices (OI; Table 2) express the variance in the TL of the consumer prey groups. The redfin culter had the most diverse diets in Lake Dianchi, with the highest value of OI (0.272), followed by those for bastard carp, common carp, and bighead carp (0.197, 0.246, and 0.252, respectively). This was in agreement with the suggestion by Lindeman (1942) that predators with high TLs are supposed to have high OI values. However, this was not the case for icefish and shrimps. They occupied relatively high TLs but had lower OI (0.089 and 0.004, respectively), owing to a certain degree of diet specialisation. Groups at low TLs, such as herbivorous fish, zoobenthos groups, cladocerans, and rotifers, fed exclusively on prey in the first trophic level; therefore, the value of OI for these groups tended to be zero. Taken together, the findings suggested that the diet characteristics of different TLs amounted to low values of system OI (0.061). Connectance index (CI) is a measure of the observed number of food links in a system relative to the number of possible links (Gardner and Ashby, 1970). The value of CI depends on both the size of a system and diet matrices. CI in Lake Dianchi was 0.194 (Table 6), lower than the theoretical value (0.252) calculated using the empirical regression equation (Christensen and Pauly, 1993). On the basis of the parameters of OI and CI, we can assume that food webs in Lake Dianchi had less complicated structures but were more linear in system connection. 3.2. Trophic network analysis 3.2.1. Trophic flux and transfer efficiency The ecosystem of Lake Dianchi was found to be detrital-based since 77.5% of the total flow originating from TLI was attributed

to detritus, whereas 22.5% from primary producers. The ratio of detrivory and herbivory ratio (D:H) was 3.4:1 (Fig. 4). On the other hand, throughput in TLII was larger than in any other TLs if detritus was not taken into account in TLI (Table 5). Noticeably, the large fraction of flow in TLII cycled to detritus, up to 42,816 t km−2 year−1 (71.60% of throughput in TLII). Cladocerans were found to be key players in this process since it required almost 69.11% (41,330 t km−2 year−1 ) of throughput in TLII, and nearly 66.73% (28,574.54 t km−2 year−1 ) of TLII-to-detritus flows. With regard to the transfer efficiencies (TEs) in the food chain, the values from TLII to TLVI were 2.9%, 6.8%, 6.1%, 7.2%, and 8.4%, respectively. The average values were calculated to be 5.1% for the grazing food chain and 4.9% for the detrital food chain (Table 5). The overall average TEs (4.9%) were about half of the assumed value (10%) (Lindeman, 1942). This low TE value might be attributed to the vital role of detritus in the food chain. Almunia et al. (1999) compared three successive stages of predominance of primary producers in Maspalomas Lagoon and found that the highest value of detritivory coincided with the lowest values of TE. 3.2.2. Ecosystem properties Zooplanktivore fish owned a large proportion of fishery; the mean trophic level of the catch was thus calculated to be 3.06 (Table 6), which was considerably similar to the value for TLIII. It was slightly higher than the value (2.92) for Lake Taihu (Li et al., 2009a,b), and similar to the value (3.07) for Northern and Central Adriatic Sea (Coll et al., 2007). Clearly, these two ecosystems have experienced both high fishing pressure and eutrophication. The total system throughput (TST; 144,460.766 t km−2 y−1 ) consisted of consumption (42.45%), flows into detritus (39.23%), respiratory flows (10.85%), and exports (7.21%). Extremely high value of TST was attributed to an increased activity of the system during cyanobacterial blooms. The ratio of compartment production to biomass (P/B) was 61.812 year−1 in Lake Dianchi. This figure

Fig. 4. Aggregation of flows and biomass in Lake Dianchi ecosystem into discrete trophic levels sensu Lindeman. D represents detritus, which has been separated to reflect the significance of recycling of non-living material through the chain. P represents primary producers, including phytoplankton and macrophytes.

K. Shan et al. / Ecological Modelling 291 (2014) 82–95 Table 6 Summary statistics of the indices estimated for the Dianchi Ecopath model. Parameter Sum of all consumption (t km−2 y−1 ) Sum of all exports (t km−2 y−1 ) Sum of all respiratory flows (t km−2 y−1 ) Sum of all flows into detritus (t km−2 y−1 ) Total system throughput (t km−2 y−1 ) Sum of all production (t km−2 y−1 ) Mean trophic level of the catch Gross efficiency (catch/net primary production) Net primary production (t km−2 y−1 ) Total primary production/total respiration Net system production (t km−2 y−1 ) Total primary production/total biomass Total biomass/total throughput Total biomass (excluding detritus) (t km−2 ) Primary production required/total primary production (%) Connectance index System omnivory index Throughput cycled (excluding detritus) Throughput cycled (including detritus) Finn’s cycling index Comprehensive cycling index (estimated by FCI) Finn’s mean path length Finn’s straight-through path length (without detritus) Finn’s straight-through path length (with detritus) Ascendancy (%) Ascendancy (flowbits) Average mutual information (AMI)

Value 61,614.648 10,418.098 15,675.121 56,752.813 144,460.766 30,578.686 3.064 0.001 26,093.305 1.665 10,418.093 61.812 0.003 422.141 53.58 0.194 0.061 0.000 57,752.230 39.980 45.657 5.536 1.972 3.323 25.300 134,250.700 0.929

was considerably larger than that of the eutrophic Lake Taihu in the 1990s (11.66 year−1 ) and Lake Nokoné (23.78 year−1 ; Li et al., 2009a,b; Villanueva et al., 2006). Brando et al. (2004) suggested that, the higher the TST value, the more rapid is the turnover rate of biomass of the entire ecosystem. Barausse et al. (2009) also suggested that a positive correlation existed between P/B and eutrophicated state of an ecosystem. According to Odum (1969), the ratio between primary production and respiration (P/R) would decline to 1 when an ecosystem becomes ‘mature’. As shown in Table 6, the P/R was 1.665 for Lake Dianchi. This result seems to explain well the eutrophic status of Lake Dianchi.

89

and was thought to have used recycled detritus to maintain its resilience. The ascendency (A), the product of TST and the average mutual information (AMI), is the key index that characterises the degree of system development and organisation. Ecosystems with higher values of ascendency (%) indeed reflect relatively higher levels of maturity. Further, ascendency as an ecological indicator could reflect the gradient of eutrophication (Patrício et al., 2004). Eutrophic processes would decrease the information diversity in an ecosystem and lead to relatively low level of ascendency. In this study, because of the low value of AMI (0.929) in Lake Dianchi, the value of ascendency (25.3%) was similar to that of the eutrophic Lake Taihu (25.7%) and Northern Adriatic Sea (24.7%) (Li et al., 2009a,b; Coll et al., 2007), and was considerably lesser than that of the relatively oligotrophic ecosystems such as Lake Qiandao (33.4%) and Évbrié lagoon (34.0%) (Villanueva et al., 2006; Liu et al., 2007). 3.2.4. Mixed trophic impact Ulanowicz and Puccia (1990) developed a mixed trophic impact (MTI) routine that has been implemented in the Ecopath system to assess the direct and indirect interactions of biomass variations of a group with any other group. The MTI analysis in Lake Dianchi highlighted that an increased abundance of primary producers, detritus, and copepods would result in positively cumulative effects on other groups, whereas ominivorous and zooplanktivorous fish, shrimps, zoobenthos groups, cladocerans, and rotifers would generate different levels of negative impacts on the ecosystem. In addition, biomass variation of filter-feeding fish (bighead and silver carps) exerted little impacts on other groups (Fig. 5). The increase of cladoceran biomass was found to have strong negative effects on most of the other groups; on the other hand, increase in the detritus biomass seemed to be beneficial to nearly all the groups. The result clearly showed that Lake Dianchi was a bottom-up functional group of the ecosystem. Interestingly, the increase of biomass of redfin culter had positive effects on the icefish (Fig. 6). The food resource competition between icefish and shrimps was so severe that the increase of redfin culter could directly prey on shrimp and alleviate this competition. 4. Discussion

3.2.3. Cycling and information indices Given the finite amount of nutrients and trace elements in an ecosystem, the same material is widely accepted to be utilised repeatedly by different organisms (Allesina, 2009). Cycling of nutrients and flows within ecosystems was believed to inflate the energy to the various end compartments. Although the network-building process (Abarca-Arenas and Ulanowicz, 2002) and computational method (Allesina and Ulanowicz, 2004) are likely to determine the quantification of cycling, Finn’s cycling index (FCI; Finn, 1976) was widely used to calculate the flows with respect to cycled fractions in an ecosystem. The FCI value in Lake Dianchi was as high as 39.98%, which was in contrast with the obtained values for both eutrophic ecosystems – Lake Taihu (11.58%) and Northern Adriatic Sea (24.85%), as well as for the macrophyte-dominated shallow Lake Gehu (14.76%) and Lake Bao’an (9.25%; Barausse et al., 2009; Jia et al., 2012; Guo et al., 2013). Cycling is assumed to increase as systems mature and is known to be related to system stability (Odum, 1969). This is because low cycling is highly dependent on rapidly energy passing through and is rather unstable and vulnerable to the changes in nutrient input (Christensen and Pauly, 1993). Moreover, a high cycling flow could also actually be a sign of stress, especially if most of the cycling occurs over short periods near the base of the trophic ladder (Christian et al., 2005). Thus, Lake Dianchi was found to have suffered from extensive stress from external environment changes

One of the most significant characteristics of the ecosystem in Lake Dianchi was a less complicated food web with extremely large throughput in the lower TLs. The ecosystem structure in YunnanGuizhou plateau was different from most lakes in the Yangtze River basin where the ecosystem showed more diversity of fish community (almost 40–70 species) and a large proportion of piscivorous fish (Liu, 1984). In addition, the historical loss of biodiversity in Lake Dianchi was attributed to the effect from deterioration of the ecosystem triggered by eutrophication (Jeppesen et al., 2000) and introduction of exotic fish (Didham et al., 2005; Puth and Post, 2005). Conventionally, it has been stressed that ecosystem are predominantly controlled by self-correcting negative feedback mechanism. The typical case was the concept of homeostasis in ecology (Morgan Ernest and Brown, 2001). However, whether the feedback was positive or negative depends completely on the relative reference frame. It was well known that many ecological processes have been considered stabilised by negative feedback. If viewed from another reference frame, they may equally well be situations in which positive feedback features prominently (Stone and Weisburd, 1992; Stone and Berman, 1993). In this study, trophic flows were chosen as reference frame. Therefore, we defined the recycling to be a loop of positive feedback in the food web of Lake Dianchi. High cycling flows concentrated in the microbial loop that would lock

90

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

Sum of MTI value

4 3 2 1 0 -1 -2 -3 -4

Fig. 5. Cumulative effects of the mixed trophic impacts (MTI) value of each impacting group on other groups.

the nutrients into the base of the food web, making nutrients available for phytoplankton uptake many times over. As a result, this positive feedback features promotes extremely high biomass in plankton communities, especially for maintaining high biomass standing stocks of cyanobacteria and primary production in Lake Dianchi. The similar stimulation of positive feedback mechanism was also found for streams and aquaculture ponds (Sterner, 1986). A consensus on the pattern of trophic TEs in the food web is that the lower the TLs, the higher the values of TE (Walline et al., 1993). However, this was not the case for Lake Dianchi. At TLII, the value of TE was extremely low owing to the bulged zooplankton biomass, especially from cladocerans. Moreover, at the upper TLs, TE values were also found to be lower than the assumed values (10%) (Lindeman, 1942). This was largely due to the lack of equal proportion of piscivorous fish. Thus, low TE values in the food web

in Dianchi indicate that there was a short cut and flow loop on the whole scale of food chain, leading to a considerable proportion of flows directly back to detritus. Overall, Lake Dianchi was found to be a detritus-driven ecosystem. Generally, detritus is known to affect energy and nutrient TEs across trophic levels and enhance food web stability. High detrivory in a food chain might mitigate resource limitations caused by environmental changes (Moore et al., 2004). According to Hairston and Hairston (1993), detritus impinging on the trophic structure could even support larger predator biomass and longer food chains. This fact was verified by Finn’s mean path length in Lake Dianchi. The path length would increase from 1.972 to 5.536 if detritus were incorporated into the calculation, indicating the obvious diet shift from predatory-type to detritivory-type. On the other hand, the role of detritus was emphasised in material storage and cycling (Cross

Negative Redfin culter Common carp Bastard carp Bighead carp Silver carp Ice fish Herbivorous fishes Small fishes Shrimps Gastropods Chironomid larvae Oligochaetes Copepods Cladocerans Rotifers Protozoa Macrophytes Phytoplankton Detritus Fleet1

Impacting group

Fleet1

Positive Detritus

Phytoplankton

Macrophytes

Protozoa

Rotifers

Cladocerans

Copepods

Oligochaetes

Chironomid larvae

Gastropods

Shrimps

Small fishes

Herbivorous fishes

Ice fish

Silver carp

Bighead carp

Bastard carp

Common carp

Redfin culter

Impacted group

Fig. 6. The direct and indirect interactions among groups in Lake Dianchi assessed using mixed trophic impacts (MTI). Positive impacts are shown above the baseline (white bar) and negative are shown below (black bar).

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

et al., 2007). It provided an ample source of organic carbon and nutrients for the organisms of the microbial loop (Karjalainen et al., 2007). High cycling flows concentrated in the microbial loop might be a rewarding process by which an ecosystem is able to recirculate and re-mineralise the nutrients. Because of the short cut between detritus and TLII, the food web of Dianchi Lake had a typically higher FCI value, reflecting a higher re-circulation and re-mineralisation of nutrients internally. Therefore, the reason Dianchi Lake could sustain heavy cyanobacterial blooms could be partly attributed to the characteristics of the food web structure. Vasconcellos et al. (1997) had shown that a system with higher capacity to recycle detritus is able to recover from perturbations by simulating the dynamics of 18 ECOPATH marine trophic models. Thus, Lake Dianchi clearly used the recycled detritus to maintain its resilience, although it suffered from multiple stresses from external environments. A note-worthy finding is that a top-down trophic cascade (Persson, 1999) had little effect on the phytoplankton biomass in Dianchi Lake. Although a certain amount of bighead and silver carps were introduced into this lake every year, the frequency and density of cyanobacterial blooms remained high. The MTI analysis revealed that filter-feeding carps hardly affected other aquatic organisms (Fig. 6). In comparison, artificially stocked filterfeeding carps were found to be effective in inhibiting cyanobacterial blooms in Lake Donghu and Lake Qiandaohu (Xie and Liu, 2001; Liu et al., 2007). The less effectiveness of top-down trophic cascade in Dianchi Lake might be attributed to the relatively small biomass of the two carps in Dianchi (0.6 t km−2 ) compared to that of Lakes Qiandaohu (6.3 t km−2 ) and Donghu (88.12 t km−2 ) and sufficient food resources for the two carps because of the extremely high biomasses of phytoplankton and zooplankton in Lake Dianchi. Rondel et al. (2008) suggested that inedible cyanobacteria (Microcystis spp.) and excess nutrients would dampen fish effects, and community-level trophic cascades would not occur. Therefore, there existed a vacancy in ecological niche for filter-feeding fish in Lake Dianchi. There are controversial views on whether cyanobacterial blooms consequently result in a trophic dead end in a food chain (Wilson et al., 2006). The traditional views insisted that the clones of Microcystis are inedible to grazers (Wiegand and Pflugmacher, 2005; Perga et al., 2013); thus, the energy bottleneck existing between phytoplankton and zooplankton in eutrophic lakes could directly restrict the energy flow to the upper TLs (Porter and Orcutt, 1980). Recently, some studies confirmed that cyanobacteria blooms and dead organisms (detritus) could be regarded as important food sources for zooplankton consumers, which eventually supported a significant share of the secondary production of lakes (Tillmanns et al., 2008). Carbon stable-isotope approach and carbon budget method have elucidated that heterotrophic bacteria were an important link between Microcystis and zooplankton (Christoffersen et al., 1990; de Kluijver et al., 2012). These microheterotrophic organisms, including flagellates and small ciliates, feed on bacterioplankton and form a ‘microbial loop’. In this study, non-living organic matter and bacteria were compacted into a single compartment for the simplification of the modelling, hampering the recognition of the role of the microbial loop. Nevertheless, we

91

could postulate that metazoan zooplankton obtained carbon via the microbial loop as evidenced by the presence of a bulged metazoan zooplankton biomass and high predating pressure of protozoa (0.931 of EE) in our study. As for the metazoan zooplankton community structure, most of the studies attempted to elucidate the distribution pattern with gradients of geographic region and trophic status (Hwang and Heath, 1997; Pinto-Coelho et al., 2005). Small omnivorous cladocerans and copepods are widely accepted to be usually dominated in subtropical eutrophic lakes (Wang et al., 2007). Apart from the above-mentioned factors, toxic cyanobacteria might also affect herbivore zooplankton either directly or indirectly. Difference in tolerance of zooplankton species against toxic Microcystis was considered to be one of the reasons for zooplankton succession with Microcystis blooms (Guo and Xie, 2006). In Dianchi Lake, MC-containing and MC-free strains usually coexisted, but largesized Microcystis (over 100 ␮m) accounting for approximately 80% of the total biomass (unpublished data) could account for the potentially high concentration of MCs (Wang et al., 2013). The zooplankton structure accompanied by cyanobacterial blooms in Lake Dianchi was in line with the findings of Hansson et al. (2007) and Moustaka-Gouni et al. (2006). Toxic cyanobacteria were shown to be negatively correlated with large unselective herbivores such as Daphnia and calanoid copepods, whereas they were positively correlated with small, relatively inefficient phytoplankton feeders such as cyclopoid copepods, bosmina, and rotifers. The predating pressure of zooplankton groups can be reflected by the value of EE. As for cladocerans and rotifers, their EE values were 0.172 and 0.311, respectively, indicating that huge resources were included in the detritus instead of being transferred to the higher TL. The EE of copepods was as high as 0.977, implying that this group was completely utilised by predators. Low EE of cladocerans was due to the high abundance of small-sized cladocerans (such as Ceriodaphnia and Bosmina) following the occurrence of cyanobacterial blooms, which allowed the alleviation of the loss from predation. Copepods mainly consisted of small cyclopoid copepods (M. varicans) and Cyclopoid nauplii. The latter was considered to be the best food resources for shrimps and icefish (Shi, 1995). The icefish (Neosalanx taihuensis Chen) in Lake Dianchi foraged exclusively on crustacean zooplankton and preferred copepods and large cladocerans (Liu and Zhu, 1994). Successful colonisation of icefish in Lake Dianchi in 1980s could be attributed to the low diversity of native competitors (Chen et al., 2001), relative stability of environmental conditions (low Chlorophyll a values listed in Fig. 2), and availability of sufficient food resources (high abundance of copepods and large cladocerans listed in Table 7). The ecosystem provided a vacancy in the ecological niche in favour of icefish in this period. Likewise, acclimatisation of icefish has also been observed in other Yunnan plateau lakes such as Lakes Fuxian, Erhai, Xingyun, and Chenghai (Guo et al., 2009). The yield of icefish, however, declined remarkably at the beginning of the 1990s in Lake Dianchi, coincident with the mass increase of cyanobacterial blooms from 1995 (Fig. 2). The decline of the icefish was perhaps attributed to the shortage of zooplankton resources since large

Table 7 Comparison of abundance of zooplankton among the four groups in different decades of Lake Dianchi ecosystem.

Protozoa (ind./L) Rotifers (ind./L) Cladocerans (ind./L) Copepods (ind./L)

1957

1982

1994

2001

2009

1200 300 55 130

6553 4554 212 203

23,325 7052 420 92

2142 4705 615 125

9616 2053 445 112

Date were average values on the basis of different sites and seasons. The studies are as follows: 1957 and 1982 by Wang (1985); 1994 by Peng (1995); 2001 and 2009 from Lake Dianchi Field Station.

92

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

herbivorous cladocerans shifted to small omnivorous ones, as well as there was a decrease in copepods during the 1990s (Table 7). In Fuxian Lake, the presence of exotic icefish was also found to be positively correlated with small cladocerans (Bosmina) density (Liu et al., 2009; Li et al., 2011). Thus, the change in zooplankton structure was thought to be in favour of cyanobacterial growth since they would encounter less predator pressure by large herbivore cladocerans. Therefore, exotic zooplanktivore fish should be cautiously introduced in Lake Dianchi and other lakes in Yunnan Lake. 5. Conclusions This study quantitatively determined the food web structure and trophic interactions in Dianchi Lake, a heavy cyanobacterial bloom-dominated lake, by using the Ecopath model. Energy fluxes were found to be largely detritus-driven, and microbial loop was considered to be involved in linking the transfer between detritus and TLII, as reflected by the bulged metazoan zooplankton. On the other hand, Lake Dianchi was relatively stable and resilient after perturbation owing to its rapid turnover among several species, with high abundance and high cycling of fluxes in low TLs. Thus, Dianchi Lake was clearly found to be a bottom-up control ecosystem. These characteristics of the food web partly explained why the cyanobacterial bloom was exceptionally heavy and durable in this lake. The understanding of the ecosystem of Dianchi Lake, therefore, helps guide the restoration. For instance, mitigation of the biomass of cyanobacteria and major nutrients should be conducted as a priority, along with steady stocking of filter-feeding carps and piscivorous carps in order to occupy the empty niche.

The ecosystem model allows placing pathways and cycling of material flows into the context of the food web to better understand the properties of cyanobacterial bloom-dominated ecosystems. However, this study has limitations, in that all dead organisms and bacteria were included in a single functional group. In the future, we recommend splitting detritus into organic and inorganic nutrients for better understanding the recycling of nutrients. This might lead to data scarcity and model indeterminacy. This task, however, is unlikely to be possible under the framework of Ecopath software. Other quantitative methods such as linear inverse models (LIM) would be more suitable for understanding the underdetermined model. Acknowledgments This work was supported by the Major Science and Technology Program for Water Pollution Control and Treatment (2013ZX07102-005, 2009ZX07106-001-002), National Basic Research Program of China (2008CB418006) and Natural Science Foundation of China-Yunnan Project (U0833604). We would like to thank Dr. Zhicai Xie and Mr. Youqin Xu for providing data of macroinvertes and zooplankton, respectively. Special thanks also to Dianchi Lake Field Station of Institute of Hydrobiology, CAS. Appendix A. See Tables A1 and A2.

Table A1 Spatial distribution of nutrients and some ecosystem components in different regions of Lake Dianchi during 2009–2010. North TP (mg/L) TN (mg/L) Chla (␮g/L) Percentage of cyanobacteria (%) Protozoa (mg/L) Rotifers (mg/L) Cladocerans (mg/L) Copepods (mg/L) Oligochaetes (g/m2 ) Chironomid larvae (g/m2 ) Gastropods (g/m2 )

0.31 3.43 128.21 68.90 0.67 2.20 10.02 1.88 6.83 1.52 1.45

Centre ± ± ± ± ± ± ± ± ± ± ±

0.07 0.79 35.32 1.00 0.04 0.59 2.32 0.47 5.57 1.31 3.22

0.20 ± 2.25 ± 86.92 ± 61.40 ± 0.42 ± 1.25 ± 5.60 ± 1.68 ± 1.35 ± 0.16 ± –

South 0.01 0.10 4.63 3.50 0.08 0.18 1.80 0.55 1.35 0.27

0.18 ± 2.33 ± 73.67 ± 54.60 ± 0.38 ± 1.10 ± 3.14 ± 2.03 ± 0.29 ± 0.025 ± –

0.00 0.14 4.24 2.30 0.06 0.23 0.39 0.38 0.21 0.03

Table A2 The formulas of some parameters used in the article. Parameter

Formula

TST (total system throughput)

TST =



Tpq = Total export + total repiration + total flows to detritus

pq

where p and q can represent either an arbitrary system component or the environment OI (omnivory index)

OIi =

n 

2

(TLj − (TLi − 1)) DCij

j=1

CI (connectance index) TL (trophic levels)

where TLj is the trophic level of prey j, TLi is the trophic level of the predator i, and, DCij is the proportion prey j constitutes to the diet of predator i MTIi,j = DCi,j − FCj,i where DCi,j is the diet composition term expressing how much j contributes to the diet of i, and FCj,i is a host composition term giving the proportion of the predation on j that is due to i as a predator CI = N/(N − 1)2 ≈ 1/(N − 1) where N is the number of living groups TL = 1 + [the weighted average of the prey’s trophic level]

Ascendency

A=

MTI (mixed trophic impact)

 i

Tij log

 Tij T..  T.j Ti.

j

T

where a dot as a subscript indicates summation over that index Average mutual information

AMI =

ij

T..

log

 Tij T..  T.j Ti.

i,j

FCI (Finn’s cycling index) PL (Finn’s mean path length)

FCI = Tc/TST where Tc is the amount of system activity involved in cycling PL = TST/(total export + total respiration)

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

References Abarca-Arenas, L.G., Ulanowicz, R.E., 2002. The effects of taxonomic aggregation on network analysis. Ecol. Model. 149 (3), 285–296. Allen, K.R., 1971. Relation between production and biomass. J. Fish. Board Can. 28 (10), 1573–1581. Allesina, S., 2009. Cycling and cycling indices. In: Jørgensen, S.E., Fath, B.D. (Eds.), Systems Ecology. Vol. 1 of Encyclopedia of Ecology. Elsevier, Oxford, pp. 812–819. Allesina, S., Ulanowicz, R.E., 2004. Cycling in ecological networks: Finn’s index revisited. Comput. Biol. Chem. 28 (3), 227–233. Almunia, J., Basterretxea, G., Aristegui, J., Ulanowicz, R.E., 1999. Benthic-pelagic switching in a coastal subtropical lagoon. Estuar. Coast. Shelf Sci. 49 (3), 363–384. Barausse, A., Duci, A., Mazzoldi, C., Artioli, Y., Palmeri, L., 2009. Trophic network model of the Northern Adriatic Sea: analysis of an exploited and eutrophic ecosystem. Estuar. Coast. Shelf Sci. 83 (4), 577–590. Beardall, J., Raven, J.A., 2004. The potential effects of global climate change on microalgal photosynthesis, growth, and ecology. Phycologia 43 (1), 26–40. Benzie, J.A.-H., 2005. Cladocera: The Genus Daphnia (Including Daphniopsis). Backhuys Publishers, Leiden, pp. 376. Berger, C., 1989. In situ primary production, biomass and light regime in the Wolderwijd, the most stable Oscillatoria agardhii lake in The Netherlands. Hydrobiologia 185 (3), 233–244. Berlow, E.L., Neutel, A.M., Cohen, J.E., De Ruiter, P.C., Ebenman, B.O., Emmerson, M., Fox, J.M., Jansen, V.A.A., Jones, J.I., Kokkoris, G.D., Logofet, D.O., McKane, A.J., Montoya, J.M., Petchey, O., 2004. Interaction strengths in food webs: issues and opportunities. J. Anim. Ecol. 73 (3), 585–598. Beverton, R.J.H., Holt, S.J., 1957. On the dynamics of exploited fish populations. Fishery Investigations Series II, vol. XIX. Ministry of Agriculture, Fisheries and Food, pp. 22. Bird, D.F., Kalff, J., 1984. Empirical relationships between bacterial abundance and chlorophyll concentration in fresh and marine waters. Can. J. Fish. Aquat. Sci. 41 (7), 1015–1023. Brando, V.E., Ceccarelli, R., Libralato, S., Ravagnan, G., 2004. Assessment of environmental management effects in a shallow water basin using mass-balance models. Ecol. Model. 172 (2), 213–232. Capriulo, G.M., Smith, G., Troy, R., Wikfors, G.H., Pellet, J., Yarish, C., 2002. The planktonic food web structure of a temperate zone estuary, and its alteration due to eutrophication. Hydrobiologia 475/476, 263–333. Carey, C.C., Ibelings, B.W., Hoffmann, E.P., Hamilton, D.P., Brookes, J.D., 2012. Ecophysiological adaptations that favour freshwater cyanobacteria in a changing climate. Water Res. 46 (5), 1394–1407. Carpenter, S.R., Caraco, N.F., Correll, D.L., Howarth, R.W., Sharpley, A.N., Smith, V.H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 8 (3), 559–568. Carpenter, S.R., Fisher, S.G., Grimm, N.B., Kitchell, J.F., 1992. Global change and freshwater ecosystem. Annu. Rev. Ecol. Syst. 23, 119–139. Chen, Y., Yang, J., Zhou, W., Cui, G., Kuang, P., 1994. The biology of Cultrichthys erythropterus in Dianchi Lake with evaluation on its influence on Neosalanx taihuensis. Zool. Res. 15, 88–95 (in Chinese with English abstract). Chen, Z.M., Yang, J.X., Su, R.F., Chen, X.Y., 2001. Present status of the indigenous fishes in Dianchi Lake, Yunnan. Biodivers. Sci. 9 (4), 407–413 (in Chinese with English abstract). Cho, B.C., Azam, F., 1990. Biogeochemical significance of bacterial biomass in the ocean’s euphotic zone. Mar. Ecol. Prog. Ser. 63 (2), 253–259. Christensen, V., Pauly, D., 1992. A Guide to the Ecopath II Software System (Version 2.1). International Center for Living Aquatic Resources Management. Christensen, V., Pauly, D., 1993. Flow characteristics of aquatic ecosystems. In: Christensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic Ecosystems. ICLARM Conference Proceedings, vol. 26, p. 341. Christensen, V., Walters, C., Pauly, D., 2000. Ecopath with Ecosim: A User’s Guide. Fisheries Centre, University of British Columbia, Vancouver, pp. 130. Chrisenten, V., Waters, C.J., 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecol. Model. 172 (2), 109–139. Christian, R.R., Baird, D., Luczkovich, J., Johnson, J.C., Scharler, U.M., Ulanowicz, R.E., 2005. Role of network analysis in comparative ecosystem ecology of estuaries. In: Aquatic Food Webs. Oxford University Press, Oxford, pp. 25–40. Christoffersen, K., Riemann, B., Hansen, L.R., Klysner, A., Sørensen, H.B., 1990. Qualitative importance of the microbial loop and plankton community structure in a eutrophic lake during a bloom of cyanobacteria. Microb. Ecol. 20 (1), 253–272. Coll, M., Santojanni, A., Palomera, I., Tudela, S., Arneri, E., 2007. An ecological model of the Northern and Central Adriatic Sea: analysis of ecosystem structure and fishing impacts. J. Mar. Syst. 67 (1), 119–154. Darwall, W.R., Allison, E.H., Turner, G.F., Irvine, K., 2010. Lake of flies, or lake of fish? A trophic model of Lake Malawi. Ecol. Model. 221 (4), 713–727. de Kluijver, A., Yu, J., Houtekamer, M., Middelburg, J.J., Liu, Z., 2012. Cyanobacteria as carbon source for zooplankton in eutrophic Lake Taihu, China, measured by 13 C labeling and fatty acid biomarkers. Limnol. Oceanogr. 57 (4), 1245–1254. Didham, R.K., Tylianakis, J.M., Hutchison, M.A., Ewers, R.M., Gemmell, N.J., 2005. Are invasive species the drivers of ecological change? Trends Ecol. Evolut. 20 (9), 470–474. Drenner, R.W., Hambright, K.D., 2002. Piscivores, trophic cascades, and lake management. ScientificWorldJournal 2, 284–307. Dussart, B.H., Defaye, D., 2001. Introduction to the Copepoda. Backhuys. Fetahi, T., Mengistou, S., 2007. Trophic analysis of Lake Awassa (Ethiopia) using mass-balance Ecopath model. Ecol. Model. 201 (3), 398–408.

93

Fetahi, T., Schagerl, M., Mengistou, S., Libralato, S., 2011. Food web structure and trophic interactions of the tropical highland lake Hayq, Ethiopia. Ecol. Model. 222 (3), 804–813. Finn, J.T., 1976. Measures of ecosystem structure and function derived from analysis of flows. J. Theor. Biol. 56 (2), 363–380. Garcia, S.M., 2003. The Ecosystem Approach to Fisheries: Issues, Terminology, Principles, Institutional Foundations, Implementation and Outlook (No. 433). Food & Agriculture Organization, pp. 71. Gardner, M.R., Ashby, W.R., 1970. Connectance of large dynamic (cybernetic) systems: critical values for stability. Nature 228 (5273), 784. Garten, D., Adrian, R., 2002. Species-specific changes in the phenology and peak abundance of freshwater copepods in response to warm summers. Freshw. Biol. 47 (11), 2163–2173. Gong, Z.J., Li, Y.L., Shen, J., Xie, P., 2009. Diatom community succession in the recent history of a eutrophic Yunnan Plateau lake, Lake Dianchi, in subtropical China. Limnology 10 (3), 247–253. Cross, W.F., Wallace, J.B., Rosemond, A.D., 2007. Nutrient enrichment reduces constraints on material flows in a detritus-based food web. Ecology 88 (10), 2563–2575. Guo, C.B., Ye, S.W., Lek, S., Liu, J.S., Zhang, T.L., Yuan, J., Li, Z.J., 2013. The need for improved fishery management in a shallow macrophytic lake in the Yangtze River basin: evidence from the food web structure and ecosystem analysis. Ecol. Model. 267, 138–147. Guo, N., Zhang, M., Yu, Y., Qian, S.Q., Li, D.M., Kong, F.X., 2009. Crustacean zooplankton communities in 13 lakes of Yunnan-Guizhou plateau: relationship between crustacean zooplankton biomass or size structure and trophic indicators after invasion by exotic fish. Ann. Limnol. Int. J. Limnol. EDP Sci. 45 (04), 279–288. Guo, N.C., Xie, P., 2006. Development of tolerance against toxic Microcystis aeruginosa in three cladocerans and the ecological implications. Environ. Pollut. 143, 513–518. Hairston Jr., N.G., Hairston Sr., N.G., 1993. Cause-effect relationships in energy flow, trophic structure, and interspecific interactions. Am. Nat., 379–411. Halfon, E., Schito, N., Ulanowicz, R.E., 1996. Energy flow through the Lake Ontario food web: conceptual model and an attempt at mass balance. Ecol. Model. 86 (1), 1–36. Han, J.H., Kumar, H.K., Lee, J.H., Zhang, C.I., Kim, S.W., Lee, J.H., Kim, S.D., An, K.G., 2011. Integrative trophic network assessments of a lentic ecosystem by key ecological approaches of water chemistry, trophic guilds, and ecosystem health assessments along with an ECOPATH model. Ecol. Model. 222 (19), 3457–3472. Hansson, L.A., Gustafsson, S., Rengefors, K., Bomark, L., 2007. Cyanobacterial chemical warfare affects zooplankton community composition. Freshw. Biol. 52 (7), 1290–1301. Heisler, J., Glibert, P.M., Burkholder, J.M., Anderson, D.M., Cochlan, W., Dennison, W.C., Dortch, Q., Gobler, C.J., Heil, C.A., Humphries, E., Lewitus, A., Magnien, R., Marshall, H.G., Sellner, K., Stockwell, D.A., Stoecker, D.K., Suddleson, M., 2008. Eutrophication and harmful algal blooms: a scientific consensus. Harmful Algae 8 (1), 3–13. Hessen, D.O., Van Donk, E., Gulati, R., 2005. Seasonal seston stoichiometry: effects on zooplankton in cyanobacteria-dominated lake. J. Plankton Res. 27 (5), 449–460. Heymans, J.J., Baird, D., 2000. Network analysis of the northern Benguela ecosystem by means of NETWRK and ECOPATH. Ecol. Model. 131 (2), 97–119. Heymans, J.J., Shannon, L.J., Jarre, A., 2004. Changes in the northern Benguela ecosystem over three decades: 1970, 1980, and 1990. Ecol. Model. 172 (2), 175–195. Hou, G.X., Song, L.R., Liu, J., Xiao, B.D., Liu, Y.D., 2004. Modeling of cyanobacterial blooms in hypereutrophic Lake Dianchi, China. J. Freshw. Ecol. 19 (4), 623–629. Huang, X.F., 1999. Survey, Observation and Analysis of Lake Ecology. Standards Press of China, Beijing (in Chinese). Hwang, S.J., Heath, R.T., 1997. The distribution of protozoa across a trophic gradient, factors controlling their abundance and importance in the plankton food web. J. Plankton Res. 19 (4), 491–518. Jeppesen, E., Jensen, J.P., Søndergaard, M., Lauridsen, T., Landkildehus, F., 2000. Trophic structure, species richness and biodiversity in Danish lakes: changes along a phosphorus gradient. Freshw. Biol. 45 (2), 201–218. Jia, P., Hu, M., Hu, Z., Liu, Q., Wu, Z., 2012. Modeling trophic structure and energy flows in a typical macrophyte dominated shallow lake using the mass balanced model. Ecol. Model. 233, 26–30. Jørgensen, S.E., Nielsen, S.N., Jorgensen, S.E., 1991. Handbook of Ecological Parameter and Ecotoxicology. Elsevier, Amsterdam. Kann, J., Smith, V.H., 1999. Estimating the probability of exceeding elevated pH values critical to fish population in a hypereutrophic lake. Can. J. Fish. Aquat. Sci. 56 (12), 2262–2270. Karjalainen, M., Engström-Öst, J., Korpinen, S., Peltonen, H., Pääkkönen, J.P., Rönkkönen, S., Suikkanen, S., Viitasalo, M., 2007. Ecosystem consequences of Cyanobacteria in the Northern Baltic Sea. AMBIO J. Hum. Environ. 36 (2), 195–202. Ke, Z., Xie, P., Guo, L., 2008. Controlling factors of spring-summer phytoplankton succession in Lake Taihu (Meiliang Bay, China). Hydrobiologia 607 (1), 41–49. Kuns, M.M., Sprules, W.G., 2000. Zooplankton production in Lake Ontario: a multistrata approach. Can. J. Fish. Aquat. Sci. 57 (11), 2240–2247. Li, M., Xie, G.Q., Dai, C.R., Yu, L.X., Li, F.R., Yang, S.P., 2009a. A study of the relationship between the water body chlorophyll a and water quality factors of the offcoast of Dianchi Lake. Yunnan Geogr. Environ. Res. 21 (2), 102–106 (in Chinese). Li, Y.K., Song, B., Chen, Y., Chen, L., Yu, N., Olson, D., 2010. Changes in the trophic interactions and the community structure of Lake Taihu (China) ecosystem from the 1960 to 1990. Aquat. Ecol. 44 (2), 337–348.

94

K. Shan et al. / Ecological Modelling 291 (2014) 82–95

Li, Y.K., Chen, Y., Song, B., Olson, D., Yu, N., Chen, L.Q., 2009b. Ecosystem structure and functioning of Lake Taihu (China) and the impacts of fishing. Fish. Res. 95 (2), 309–324. Li, Y., Gong, Z., Xia, W., Shen, J., 2011. Effects of eutrophication and fish yield on the diatom community in Lake Fuxian, a deep oligotrophic lake in southwest China. Diatom Res. 26 (1), 51–56. Lindeman, R.L., 1942. The trophic dynamic aspect of ecology. Ecology 23 (4), 399–418. Liu, G., Liu, Z., Li, Y., Chen, F., Gu, B., Smoak, J.M., 2009. Effects of fish introduction and eutrophication on the cladoceran community in Lake Fuxian, a deep oligotrophic lake in southwest China. J. Paleolimnol. 42 (3), 427–435. Liu, J.K., 1999. Advanced Hydrobiology. Scinece Press, Beijing, pp. 133 (in Chinese). Liu, J.K., 1984. Lakes of the middle and lower basins of the changjiang (China). In: Taub, F.B. (Ed.), Ecosystems of the World 23 Lakes and Reservoirs. Elsevier, Amsterdam, pp. 331–355. Liu, Q.G., Chen, Y., Li, J.L., Chen, L.Q., 2007. The food web structure and ecosystem properties of a filter-feeding carps dominated deep reservoir ecosystem. Ecol. Model. 203 (3), 279–289. Liu, Y.P., 2001. Analysis on eutrophication of Dianchi Lake and its control measures. Yunnan Environ. Sci. S1, 25–27 (in Chinese). Liu, Z., Zhu, S., 1994. Food and feeding behavior of ice-fish (Neosalanx Taihuensis Chen) in Dianchi Lake. Acta Zool. Sin. 40, 253–261 (in Chinese with English abstract). Livingstone, D.M., 2003. Impact of secular climate change on the thermal structure of a large temperate central European lake. Clim. Change 57 (12), 205–225. Lu, J., Wang, H.B., Pan, M., Xia, J., Xing, W., Liu, G.H., 2012. Using sediment seed banks and historical vegetation change data to develop restoration criteria for a eutrophic lake in China. Ecol. Eng. 39, 95–103. McCarthy, M.J., Lavrentyev, P.J., Yang, L., Zhang, L., Chen, Y., Qin, B., Gardner, W.S., 2007. Nitrogen dynamics and microbial food web structure during a summer cyanobacterial bloom in a subtropical, shallow, well-mixed, eutrophic lake (Lake Taihu, China). In: Eutrophication of Shallow Lakes with Special Reference to Lake Taihu, China. Springer, The Netherlands, pp. 195–207. McDonald, C.P., Urban, N.R., Casey, C.M., 2010. Modeling historical trends in Lake Superior total nitrogen concentrations. J. Great Lakes Res. 36 (4), 715–721. Moore, J.C., Berlow, E.L., Coleman, D.C., Ruiter, P.C., Dong, Q., Hastings, A., Johnson, N.C., McCann, K.S., Melville, K., Morin, P.J., Nadelhoffer, K., Rosemond, A.D., Post, D.M., Sabo, J.L., Scow, K.M., Vanni, M.J., Wall, D.H., 2004. Detritus, trophic dynamics and biodiversity. Ecol. Lett. 7 (7), 584–600. Moreau, J., Mavuti, K., Daufresne, T., 2001. A synoptic Ecopath model of biomass flows during two different static ecological situations in Lake Nakuru (Kenya). Hydrobiologia 458 (1–3), 63–74. Morgan Ernest, S.K., Brown, J.H., 2001. Homeostasis and compensation: the role of species and resources in ecosystem stability. Ecology 82 (8), 2118–2132. Moustaka-Gouni, M., Vardaka, E., Michaloudi, E., Kormas, K.A., Tryfon, E., Mihalaton, H., Gkelis, S., Lanaras, T., 2006. Plankton food web structure in a eutrophic polymictic lake with a history of toxic cyanobacterial blooms. Limol. Oceanogr. 51 (1), 715–727. Odum, E.P., 1969. The strategy of ecosystem development. Science 164, 262–270. Paerl, H.W., Huisman, J., 2008. Blooms like it hot. Science 320 (5872), 57–58. Paerl, H.W., Huisman, J., 2009. Climate change: a catalyst for global expansion of harmful cyanobacterial blooms. Environ. Microbiol. Rep. 1 (1), 27–37. Paerl, H.W., Xu, H., McCarthy, M.J., Zhu, G., Qin, B., Li, Y., Gardner, W.S., 2011. Controlling harmful cyanobacterial blooms in a hyper-eutrophic lake (Lake Taihu, China): the need for a dual nutrient (N&P) management strategy. Water Res. 45 (5), 1973–1983. Paerl, H.W., Paul, V.J., 2012. Climate change: links to global expansion of harmful cyanobacteria. Water Res. 46 (5), 1349–1363. Palomares, M.L.D., Pauly, D., 1998. Predicting food consumption of fish populations as functions of mortality, food type, morphometrics, temperature and salinity. Marine Freshw. Res. 49 (5), 447–453. Park, R.A., 1974. A generalized model for simulating lake ecosystems. Simulation 23 (2), 33–50. Patrício, J., Ulanowicz, R., Pardal, M.A., Marques, J.C., 2004. Ascendency as an ecological indicator: a case study of estuarine pulse eutrophication. Estuar. Coast. Shelf Sci. 60 (1), 23–25. Pauly, D., 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. 39 (2), 175–192. Pauly, D., Christensen, V., Dalsgaard, J., Forese, R., Torres, F., 1998. Fishing down marine food web. Science 279 (5352), 860–863. Pauly, D., Christensen, V., Walters, C., 2000. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of fisheries. ICES J. Mar. Sci. 57 (3), 697–706. Peng, Q.Y., 1995. Study on zooplankton in Dianchi Lake. Reserv. Fish. 6, 22–26 (in Chinese). Peng, Q.Y., 2002. The shrimp in Dianchi Lake. Reserv. Fish. 22 (2), 33–34 (in Chinese). Perga, M.E., Domaizon, I., Guillard, J., Hamelet, V., Anneville, O., 2013. Are cyanobacterial blooms trophic dead ends? Oecologia 172 (2), 551–562. Persson, L., 1999. Trophic cascades: abiding heterogeneity and the trophic level concept at the end of the road. Oikos 85, 385–397. Pinto-Coelho, R., Pinel-Alloul, B., Méthot, G., Havens, K.E., 2005. Crustacean zooplankton in lakes and reservoirs of temperate and tropical regions: variation with trophic status. Can. J. Fish. Aquat. Sci. 62 (2), 348–361. Porter, K.G., Orcutt, J.D., 1980. Nutritional adequacy, manageability, and toxicity as factors that determine food quality of green and blue-green algae for Daphnia.

In: Evolution and Ecology of Zooplankton Communities. University Press of New England, pp. 268–281. Puth, L.M., Post, D.M., 2005. Studying invasion: have we missed the boat? Ecol. Lett. 8 (7), 715–721. Qu, Z.X., Li, H., 1983. Plant Community and Pollution in Lake Dianchi. Peoples Press of Yunnan, Kunming (in Chinese). Rice, J., 1995. Food web theory, marine food webs, and what climate changes may do to northern marine fish populations. In: Beamish, R.J. (Ed.), Climate Change and Northern Fish Populations, vol. 121. Canadian Special Publication of Fisheries and Aquatic Sciences, pp. 561–568. Rondel, C., Arfi, R., Corbin, D., Le Bihan, F., Ndour, E.H., Lazzaro, X., 2008. A cyanobacterial bloom prevents fish trophic cascades. Freshw. Biol. 53 (4), 637–651. Shi, W.G., 1995. Biology and feeding habit of Palaemon modestus (Heller) in Taihu Lake. J. Lake Sci. 7, 69–76 (in Chinese with English abstract). Song, B., 2004. Ecosystem dynamics of the fisheries and environment of Taihu Lake. Doctoral Dissertation of East China Normal University, Shanghai (in Chinese with English abstract). Sommaruga, R., 1995. Microbial and classical food webs: a visit to a hypertrophic lake. FEMS Microbiol. Ecol. 17 (4), 257–270. Sterner, R.W., 1986. Herbivores’ direct and indirect effects on algal populations. Science 231 (4738), 605–607. Stone, L., Berman, T., 1993. Positive feedback in aquatic ecosystem: the case of the microbial loop. Bull. Math. Biol. 55 (5), 919–936. Stone, L., Weisburd, R.S., 1992. Positive feedback in aquatic ecosystems. Trends Ecol. Evol. 7 (8), 263–267. Straile, D., 2002. North Atlantic Oscillation synchronizes food–web interactions in central European lakes. Proc. R. Soc. Lond. B: Biol. Sci. 269 (1489), 391–395. Sun, L.M., Chen, D.H., Wang, Q.X., Gao, Y.F., Dai, J.J., 2011. Spatio-temporal distribution of chlorophyll a concentration and primary productivity between winter in 2009 and spring in 2010 of Lake Dianchi. J. Shanghai Norm. Univ. (Nat. Sci.) 40 (2), 191–196 (in Chinese with English abstract). Thapanand, T., Moreau, J., Jutagate, T., Wongrat, P., Lekchonlayut, T., Meksumpun, C., Janekitkarn, S., Rodloi, A., Dulyapruk, V., Wongrat, L., 2007. Towards possible fishery management strategies in a newly impounded man-made lake in Thailand. Ecol. Model. 204 (1), 143–155. Tillmanns, A.R., Wilson, A.E., Pick, F.R., Sarnelle, O., 2008. Meta-analysis of cyanobacterial effects on zooplankton population growth rate: species-specific responses. Fundam. Appl. Limnol./Arch. Hydrobiol. 171 (4), 285–295. Ulanowicz, R.E., 1987. NETWRK4: A Package of Computer Algorithms to Analyze Ecological Flow Networks. University of Maryland, Chesapeake Bay Laboratory, Solomons, MD, USA. Ulanowicz, R.E., Puccia, C.J., 1990. Mixed trophic impacts in ecosystems. Coenoses 5 (1), 7–16. Van Oevelen, D., Van den Meersche, K., Meysman, F.J., Soetaert, K., Middelburg, J.J., Vézina, A.F., 2010. Quantifying food web flows using linear inverse models. Ecosystems 13 (1), 32–45. Vasconcellos, M., Mackinson, S., Sloman, K., Pauly, D., 1997. The stability of trophic mass-balance models of marine ecosystem: a comparative analysis. Ecol. Model. 100 (1), 125–134. Villanueva, M.C., Isumbisho, M., Kaningini, B., Moreau, J., Micha, J.C., 2008. Modeling trophic interactions in Lake Kivu: what roles do exotics play? Ecol. Model. 212 (3), 422–438. Villanueva, M.C., Lalèyè, P., Albaret, J.J., Laë, R., de Morais, L.T., Moreau, J., 2006. Comparative analysis of trophic structure and interactions of two tropical lagoons. Ecol. Model. 197 (3), 461–477. Walline, P.D., Pisanty, S., Gophen, M., Berman, T., 1993. The ecosystem of Lake Kinneret, Israel. In: Trophic Models of Aquatic Ecosystems. ICLARM Conference Proceedings 26, Manila, Philippines, pp. 124–129. Wan, N., Song, L.R., Wang, R.N., Liu, J.T., 2008. The spatio-temporal distribution of algal biomass in Dianchi Lake and its impact factors. Acta Hydrobiol. Sin. 32 (2), 184–188 (in Chinese with English abstract). Wang, C.M., Xie, Z.C., Song, L.R., Xiao, B.D., Li, G.B., 2011. Dianchi Lake macroinvertebrate community succession trends and retrogressive analysis. Zool. Res. 32, 212–221 (in Chinese with English abstract). Wang, S.B., Xie, P., Wu, S.K., Wang, H.J., 2007. Crustacean zooplankton size structure in aquaculture lakes: is larger size structure always associated with higher grazing pressure? Hydrobiologia 575 (1), 203–209. Wang, X.Y., Sun, M.J., Xie, M.J., Liu, M., Luo, L., Li, P.F., Kong, F.X., 2013. Differences in microcystin production and genotype composition among Microcystis colonies of different sizes in Lake Taihu. Water Res. 47, 5659–5669. Wang, Y., Wang, Y., Wang, C., Peng, Q.Y., Chen, Y.L., Yang, J.X., 1995. Study on the correlation between the growing characters of Neosalanx taihuensis and its resources in Lake Dianchi. Zool. Res. 19, 290–294 (in Chinese with English abstract). Wang, Y.C., (Doctoral Dissertation) 2010. Studys on lake ecosystems regime shift course and phase division. Institute of Hydrobiology, Chinese Academy of Sciences (in Chinese with English abstract). Wang, Z.Z., 1985. The basic survey of the zooplankton in Yunnan Dianchi Lake. J. Yunnan Univ. 7, 53–72 (in Chinese with English abstract). Wiegand, C., Pflugmacher, S., 2005. Ecotoxicological effects of selected cyanobacterial secondary metabolites a short review. Toxicol. Appl. Pharmacol. 203 (3), 201–218. Wilcove, D.S., Rothstein, D., Dubow, J., Phillips, A., Losos, E., 1998. Quantifying threats to imperiled species in the United States. Bioscience 48 (8), 607–615. Wilhelm, S.W., Farnsley, S.E., LeCleir, G.R., Layton, A.C., Satchwell, M.F., DeBruyn, J.M., Boyer, G.L., Zhu, G.W., Paerl, H.W., 2011. The relationships between nutrients,

K. Shan et al. / Ecological Modelling 291 (2014) 82–95 cyanobacterial toxins and the microbial community in Taihu (Lake Tai), China. Harmful Algae 10 (2), 207–215. Wilson, A.E., Sarnelle, O., Tillmanns, A.R., 2006. Effects of cyanobacterial toxicity and morphology on the population growth of freshwater zooplankton: metaanalyses of laboratory experiments. Limnol. Oceanogr. 51 (4), 1915–1924. Wrona, F.J., Prowse, T.D., Reist, J.D., Hobbie, J.E., Lévesque, L.M., Vincent, W.F., 2006. Climate change effects on aquatic biota, ecosystem structure and function. AMBIO J. Hum. Environ. 35 (7), 359–369. Xiao, Y., Gan, N.Q., Liu, J., Zheng, L.L., Song, L.R., 2012. Heterogeneity of buoyancy in response to light between two buoyant types of cyanobacterium Microcystis. Hydrobiologia 679 (1), 297–311. Xie, P., Liu, J., 2001. Practical success of biomanipulation using filter-feeding fish to control cyanobacteria blooms: a synthesis of decades of research and application in a subtrophical hypereutrophic lake. ScientificWorldJournal 1, 337–356. Xing, P., Kong, F.X., Cao, H.S., Zhang, M., 2007. Relationship between bacterioplankton and phytoplankton community dynamics during late spring and early summer in Lake Taihu, China. Acta Ecol. Sin. 27 (5), 1696–1702 (in Chinese with English abstract).

95

Yan, Y., Liang, Y., 2003. Energy flow of macrozoobenthic community in a macrophytic lake, Biandantang Lake. Acta Ecol. Sin. 23 (3), 527–538 (in Chinese with English abstract). Ye, S.W., (Doctoral Dissertation) 2007. Studies on fish communities and trophic network model of shallow lakes along the middle reach of the Yangtze River. Institute of Hydrobiology, Chinese Academy of Sciences (in Chinese with English abstract). Zhang, J., He, Z., 1991. Handbook of Investigation on the Fishery and Natural Resources of Inland Waters. Agriculture Press, Beijing (in Chinese). Zhang, X., Xie, P., Chen, F.Z., Li, Y.L., Li, S.X., Guo, N.C., Qin, J.H., 2005. Present status and changes of the phytoplankton community after invasion of Neosalanx taihuensis since 1982 in a deep oligotrophic plateau lake, Lake Fuxian in the subtropical China. J. Environ. Sci. 17 (3), 389–394. Zhou, D., Ye, J., Wang, W., Liu, Z., 2011. Feeding habits and the relation to the size of Cultrichthys erythropterus in Meiliang Bay, Lake Taihu: implication for biomanipulation and management. J. Lake Sci. 23 (5), 796–800. Zhuang, Y.L., Feng, Z.H., Li, J.H., 1996. Study on proliferation of Neosalanx tangkahkeii taihuensis Chen in Yunnan plateau lakes. Reserv. Fish. 3, 16–30 (in Chinese).