Elsevier

Ecological Modelling

Volume 348, 24 March 2017, Pages 14-24
Ecological Modelling

Predicting the impact of Lake Biomanipulation based on food-web modeling—Lake Kinneret as a case study

https://doi.org/10.1016/j.ecolmodel.2016.12.019Get rights and content

Abstract

Biomanipulation is a tool decision makers use to achieve desirable management goals. In lakes, one of the most common goals is the improvement of water quality, an objective that can be achieved mainly by reducing the amount of phytoplankton in the water. Although it is a very clear goal that is achievable by using actions that affect the phytoplankton biomass, experience shows that primary biomanipulation goals are rarely achieved. A biomanipulation program was conducted in Lake Kinneret over a 12-year period with the goal of improving water quality by reducing the population of the dominant fish species in the lake. However, the biomanipulation failed to achieve the goal and the program was stopped. We used Ecopath with Ecosim (EwE) scenarios to examine the effect of biomanipulation on the ecosystem. The results of these scenarios show that biomanipulation actions, such as those used in the lake, indeed fail to improve water quality; furthermore, they will actually increase the amount of phytoplankton in the water and decrease water quality. The development of the method described in the present article provides managers with the means to evaluate the effect of biomanipulation on an ecosystem. This method enables researches to conduct a pre-action analysis of the planned measures and examine whether the goal can be achieved, saving money and time and preventing damage to the ecosystem.

Introduction

Over the past 30 years it has been recognized that stocking or removing certain fishes from lakes (i.e., biomanipulation) can have an important influence on the food-web structure and the resulting water quality. Biomanipulation is increasingly used as a lake restoration technique, largely because the ability to enhance water quality and support fish populations is important to a variety of lake users (Mehner et al., 2002, Arlinghaus et al., 2015). Biomanipulation in aquatic ecosystems involves the manipulation of fish populations for the purpose of inducing a consumer-mediated trophic cascade in the food-web that will in turn influence water quality (Jeppesen et al., 2012). This technique can be integrated with other biomanipulation approaches, such as introducing or enhancing the herbivorous fish biomass in order to control the macrophytic biomass (Rowe and Champion, 1994, Yu et al., 2016).

Although biomanipulation research has contributed substantially to our understanding of lake food-webs, the successful application of these techniques is not a “one size fits all” approach. In a review of biomanipulation research, Mehner et al. (2002) discussed the factors that affect food-web complexity and the success of various biomanipulation efforts. They made the following observations: (1) nutrient recycling by aquatic organisms contributes to bottom-up impacts on lake productivity, but the magnitude of these impacts varies greatly between lakes; (2) the complexity of food-web interactions is enhanced by size-dependent interactions and bottom-up impacts on lake productivity, but the magnitude of these impacts varies greatly between fishes and can limit our ability to predict the outcome of a biomanipulation event successfully; (3) it is important to consider the temporal and spatial scales of biomanipulation research – repeated interventions may be necessary to maintain the desired outcome in the lakes; and (4) a correct balance between piscivorous, planktivorous, and benthivorous fishes is needed in order to achieve the desired biomanipulation outcome, but it can be a challenge due to a general lack of quantitative assessment. When practical, the removal of undesirable fishes seems to have a larger impact on water quality than the stocking of piscivorous fishes. A synthesis of biomanipulation studies involving 39 lakes that varied in size, between 0.18 and 2,650 ha and with a mean depth of 23 m, showed that changes in phytoplankton biomass and water transparency as a result of biomanipulation were most successful in small, shallow lakes smaller than 25 ha and with a mean depth of <3 m (Drenner and Hambright, 1999). Changes in water quality were also influenced by the type of biomanipulation implemented. Approximately 90% of the studies that implemented partial fish removal succeeded in improving water quality. There are other approaches used, with varying levels of success: (1) elimination and restocking of fish (67%), (2) partial fish removal together with piscivore stocking (60%), (3) elimination of fish (40%) and (4) piscivore stocking (27%). Still, approximately 15% of the studies that used biomanipulation techniques were unsuccessful in enhancing water quality for at least 1 year (Drenner and Hambright, 1999). Reducing the number of benthivorous fishes in the lake can also indirectly reduce algal biomass, since it triggers a shift to the clear, macrophyte-dominated state characteristic in many shallow lakes (Hubert and Quist, 2010). Yet, not all the studies that used biomanipulation met with success or produced stable ecosystems. Burns and Schallenberg (2013) summarized 50 years of biomanipulation attempts in several lakes in New Zealand and concluded that in some lakes biomanipulation has indeed resulted in a better water quality in the short term (<5 years). However, in the long-term the results of biomanipulation must be accompanied by reductions in nutrient loading, achievable only via an integrated management program that will consider both the direct and the indirect impacts of each step of a proposed biomanipulation program on the whole lake ecosystem.

A long term biomanipulation program was implemented in Lake Kinneret, Israel, between 1994 and 2006, based on the trophic cascade hypothesis and with the aim of preventing deterioration and improving water quality. The plan aimed to decrease the number of Mirogrex terraesanctae (locally known as Lavnun), with the final goal of both decreasing the predation pressure on the zooplankton and making it possible for the remaining fishes to “benefit” from more food, thus enabling them to reach commercial size in a better physical state. Beyond that, reducing the predation pressure of the Lavnun on the zooplankton would revive and increase the zooplankton population and thus allow them to graze on more phytoplankton. This, it was assumed, would improve the quality of the water in Lake Kinneret (Gasith and Zohary, 2006). In the framework of the program, quantities of between 300 and 900 tons of Lavnun were removed from the lake annually between 1994 and 2006. Most of them were not used commercially, but were buried in nearby landfills after their removal from the lake. Despite the ongoing removal program, another collapse of the Lavnun’s commercial fishing occurred in 2004/5, and the fish caught were of sub-commercial size (Hambright, 2008). The unsuccessful biomanipulation of Lake Kinneret as well as the results of biomanipulation programs conducted in different lakes in New Zealand, and elsewhere, emphasize the need for a tool that will enable managers to determine the likely outcomes of biomanipulation on the ecosystem.

The dynamic simulation modeling tool Ecosim has the ability to simulate future management scenarios and analyze the impact of different variables on the ecosystem. Ecosim, a component of the Ecopath with Ecosim software package (www.ecopath.org), expands Ecopath’s capabilities and allows the exploration of temporal impacts of fishing and environmental factors. It enables users to change fishing mortality or fishing effort over time, enabling the exploration of fishing options and changes in ecosystem functioning (Coll et al., 2015). It also dynamically responds to changes in fishing mortality and biomass, enabling the creation of dynamic simulations at the ecosystem level from the primary parameters of a baseline Ecopath model (Walters et al., 1997, Christensen et al., 2000).

Christensen and Walters (2004) used Ecosim to search for alternative exploitation patterns, setting different sustainability objectives and optimizing for profit, value and conservation in the Gulf of Thailand. Coll et al. (2009) summarized several cases where an Ecosim model was used to analyze policy optimization and management; most of the scenarios focused mainly on the aspect of fisheries management. Heymans et al. (2009) used the time dynamic model to explored alternative policy options for the northern Benguela ecosystem. Heymans et al. (2016) summarized the use of models in several aspects including management context, specifically using the concept of ‘key runs’ for ecosystem-based management.

In the present study, we used Ecopath with Ecosim (EwE) to analyze the effect the biomanipulation program had on Lake Kinneret’s ecosystem. We used an Ecopath model that was previously constructed to study the lake’s food-web (Ofir et al., 2016). In the current study we developed a time-dynamic model (Ecosim) based on the Ecopath model, which was calibrated using lake data for the time-period of 1996–2012. We compared simulations that included or excluded biomanipulation strategies in order to evaluate the impact of the biomanipulation on the ecosystem. This was done by running and comparing 10-year simulations.

Section snippets

Study area

Lake Kinneret (or, alternatively, the Sea of Galilee or Lake Tiberias), the largest freshwater body in the Middle East, is a mono-mictic subtropical lake located at ∼210 m altitude, i.e., below mean sea level. It has a surface area of approximately 167 km2 and a watershed of 2730 km2. The main inflow is from the Jordan River, which contributes on average 70% of the total inflow (Gal et al., 2003), while the main outflow, until recently (2014), was pumping to Israel’s National Water Carrier. The

Results

A model of the lake’s food-web was created as part of the output from Ecopath, defining the trophic level of all the components and the linkages between all the functional groups (Fig. 1). The Lavnun fish, group no. 14 in the model, has one of the highest trophic levels, with a large biomass compared to the rest of the fish groups. As a consequence, removing the Lavnun could potentially have a big impact on the lake’s ecosystem.

Discussion

Biomanipulation is a tool that decision makers use in order to achieve specific goals in ecosystem management. A biomanipulation program was conducted in Lake Kinneret (Israel) between 1996 and 2006 in which almost 1000 T of Lavnun were removed from the lake each year. It was, however, difficult to determine whether the program’s goals were achieved, and the program was therefore stopped in 2006. In order to demonstrate the ability of food-web models to help in the planning process of

Conclusions

In many lakes and water resources biomanipulation actions are performed in order to achieve particular goals that will also increase the benefits that can be obtained from ecosystem services. However, in many cases the results of these actions do not meet the goals, and unexpected results can occur in the ecosystem, leading to the reevaluation or even the discontinuation of the program and culminating in a loss of time and money.

The results of the present research emphasize the importance of

Acknowledgements

We thank the Kinneret Limnological Laboratory staff and researchers for sharing and providing their data. We thank Yifat Artzi for the Cormorant data. This research was supported by grants from the Israel Water Authority and the Israeli Ministry of Agriculture and Rural Development.

References (44)

  • R.N. Ahrens et al.

    Foraging arena theory

    Fish Fish.

    (2012)
  • H. Akaike

    A new look at the statistical model identification

    IEEE Trans. Autom. Control

    (1974)
  • R. Arlinghaus et al.

    Management of freshwater fisheries: addressing habitat, people and fishes

    Freshwater Fish. Ecol.

    (2015)
  • C.W. Burns et al.

    Potential use of classical biomanipulation to improve water quality in New Zealand lakes: a re-evaluation

    N. Z. J. Mar. Freshwater Res.

    (2013)
  • V. Christensen et al.

    Trade-offs in ecosystem-scale optimization of fisheries management policies

    Bull. Mar. Sci.

    (2004)
  • V. Christensen et al.

    Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of fisheries

    ICES J. Mar. Sci.

    (2000)
  • M. Coll et al.

    Ecosystem modelling using the Ecopath with Ecosim approach

  • M. Coll et al.

    Modelling dynamic ecosystems: venturing beyond boundaries with the Ecopath approach

    Rev. Fish Biol. Fish.

    (2015)
  • R.W. Drenner et al.

    Biomanipulation of fish assemblages as a lake restoration technique

    Arch. Hydrobiol.

    (1999)
  • G. Gal et al.

    Modeling the Kinneret ecosystem

  • A. Gasith et al.

    Workshop on the Issue of Lavnun Removal: Summary and Recommendations to the Water Commission and Fisheries Department (in Hebrew)

    (2006)
  • K.D. Hambright

    Long-term zooplankton body size and species changes in a subtropical lake: implications for lake management

    Fundam. Appl. Limnol.

    (2008)
  • Cited by (26)

    • Piscivore stocking significantly suppresses small fish but does not facilitate a clear-water state in subtropical shallow mesocosms: A biomanipulation experiment

      2022, Science of the Total Environment
      Citation Excerpt :

      In most studies, the effects of stocking of piscivores and potential mechanisms have been well studied in shallow lakes in temperate regions such as North Europe and North America (Jacobsen et al., 1997; Hansson et al., 1998; Carpenter et al., 2001; Skov et al., 2002; Olin et al., 2006; Potthoff et al., 2008). Despite recent advances (Mazzeo et al., 2010; Jeppesen et al., 2012; Ofir et al., 2017; Liu et al., 2018), the applicability of piscivore stocking for restoration of warm (i.e., subtropical, tropical and Mediterranean) shallow eutrophic lakes is not well assessed. Compared to temperate regions, trophic cascades are expected to occur less frequently in warm lakes due to the more complex food webs characterized by factors such as high abundance of invertebrate predators, high dominance of small-sized zooplankton species, and a high degree of fish omnivory (Lazzaro, 1997; Meerhoff et al., 2007; Jeppesen et al., 2012; Su et al., 2021).

    View all citing articles on Scopus
    View full text