Elsevier

Environmental Modelling & Software

Volume 86, December 2016, Pages 158-167
Environmental Modelling & Software

Exploring Lake Victoria ecosystem functioning using the Atlantis modeling framework

https://doi.org/10.1016/j.envsoft.2016.09.019Get rights and content

Highlights

  • Lake Victoria ecosystem functioning since 1958 is described.

  • The importance of predator-prey interactions is highlighted.

  • Introduced species show negative correlation with native species.

  • The developed model can be used to “road test” fisheries management regimes.

Abstract

Lake Victoria has experienced human induced pressures such as overfishing, introduction of alien species, increased eutrophication and climate change impacts. However, there is limited understanding of the system dynamics, major processes, drivers and responses to the changes. To address this challenge, we developed the first end-to-end whole ecosystem model (Atlantis) for the lake. The model is spatially resolved into 12 unique dynamic areas based on depth, species composition, physical-chemical characteristics and fisheries management zones. A total of 38 functional groups constitute the biological model. Four fishing fleets with different targeting options are simulated. Reliability of the model is confirmed by the good fit of simulations output to observational data sets. Herein, we describe the evolution of the biophysical system, illustrating how it responded to the aforementioned induced perturbations since 1958. The constructed virtual Lake Victoria ecosystem model provides a platform for exploring the impact of management interventions before actual implementation.

Introduction

Lake Victoria, with a surface area of 68,800 km2, is the largest tropical and second largest freshwater lake on the planet. The lake's waters straddles three countries (Kenya, Uganda and Tanzania) and supports Africa's largest inland fishery (Geheb et al., 2007). The lake is large enough to create its own weather system as well as influence regional climate (Crul, 1995, Stager and Johnson, 2007). The lake provides ecosystem services such as water for domestic and industrial use, transport, hydro-power generation and food to about 40 million people. Its fisheries provide employment, income and export earnings to the lake-edge communities. Lake Victoria is home to diverse flora and fauna which are intricately connected ecologically.

The lake's ecological health is in jeopardy, and had been for decades, mainly due to a myriad of anthropogenic activities (Hecky et al., 2010). Many riparian towns release raw sewage and municipal waste into the lake on a daily basis. This, together with fertilizer and chemicals from agricultural farms in the catchment, contribute to increased pollution and eutrophication. Invasive species such as water hyacinth (Eichhornia crassipes) and Nile perch (Lates niloticus) are thought to have been responsible for the manifest of ecological damage in Lake Victoria (Opande et al., 2004, Taabu-Munyaho et al., 2016, Witte et al., 1992). Water hyacinth reproduces rapidly and covers large areas of the lake forming dense mats of plants that block sunlight needed for survival of life below the surface. The introduction of the Nile perch had a major impact on haplochromine cichlids stocks which remain favorite prey, affecting both their abundance and diversity (Witte et al., 1992). Several haplochromine species had gone extinct and their abundance was reduced to less than 1% of their original biomass barely two decades after the introduction of Nile perch. The number of species in demise has only continued to grow and now up to 65% of haplochromine species are thought to have been lost, an event which may well represent the largest species extinction amongst vertebrates in the 20th century (Goldschmidt et al., 1993, Kaufman, 1992, Kaufman and Cohen, 2013, Kitchell et al., 1997). Another challenge is the booming fishing industry that evolved around the explosion in biomass of the Nile perch. The lake region is among the most densely populated areas in Africa (Ewald et al., 2004) and demand for fish locally has been increasing rapidly with population growth. This has lead to overexploitation of fish populations, reducing them to dangerously low levels (Taabu-Munyaho et al., 2016).

The complex ecology mixed with adverse human actions on the lake and its catchment (often without prior research of potential impacts) have limited the understanding of the system dynamics, major processes, drivers and responses with no scientific consensus on the subject (Cornelissen et al., 2015, Downing et al., 2014, Odada et al., 2009). Little headway has been made in identifying the major drivers of ecosystem changes that have been witnessed in the last six decades. Work done on the lake usually focuses on one or a few aspects of the system and often falls short of giving “the big picture”. With such a scenario in place, it is quite difficult to predict the implications, both for the ecosystem and the local communities, of any management measures instituted on the lake. In the recent past it has been acknowledged that an ecosystem approach to management is necessary if the lake is to offer ecosystem services in a sustainable way (Cornelissen et al., 2015, Downing et al., 2014, Njiru et al., 2014). This study seeks to describe the Lake Victoria ecosystem functioning, trophic cascade mechanisms as well as complex non-linear system responses. A coupled component modeling approach is employed aiming to describe complex interactions among detailed processes for the purposes of prediction, forecasting and system understanding (Bennett et al., 2013, Kelly et al., 2013). We implement these by developing the first end-to-end (whole of ecosystem) model for Lake Victoria. It is envisaged that the developed model (Atlantis) will be used to predict how the lake might respond to different management measures.

Section snippets

Materials and methods

We used the Atlantis ecosystem modelling framework to develop the model. Atlantis provides an opportunity to build a virtual ecosystem which can be used to road test different management regimes before actual implementation (Fulton et al., 2011, Smith et al., 2015). In its fullest form it considers all aspects (parts) of an ecosystem i.e. biophysical, economic and social. The model tracks changes in three dimensional space consisting of horizontal polygons and vertical layers. This 3D structure

Results

The developed Atlantis model simulates the Lake Victoria ecosystem functioning from 1958 to 2015. The model runs for 57 years including 10 years burn-in for the system to stabilize. The model's performance is assessed by comparing the general patterns of primary production (Chlorophyll a), fish biomass and time series catch trends with observational data. Fig. 4 shows the spatial distribution of nitrates in the model which is required for the growth of primary producers. Higher concentrations

Discussion

This study presents the first end-to-end ecosystem model ever for Lake Victoria. Physical-chemical processes, interaction of biological groups with the environment and impact of fishing fleets are simulated. Model validity is affirmed by its ability to produce distributions of nutrients, primary production and major fish species that generally match with observed data and those reported in the literature. The temporal trends of biomass of fish species match well with the variation of reported

Conclusion

Lake Victoria is vast and of immense ecological and socio-economic significance for the riparian communities. However, sustainable management of the lake has been a challenge because of inadequate information about the resource dynamics. This study developed an Atlantis ecosystem model to simulate the physical, chemical and biological processes and how the ecosystem responds to anthropogenic activities. Reliability of the model is supported by the ability of the simulations to match well with

Software availability

Atlantis: Model documentation found on https://confluence.csiro.au/display/Atlantis/Atlantis+Ecosystem+Model+Home+Page, Author: Elizabeth A. Fulton CSIRO Wealth from Oceans Flagship, Division of Marine and Atmospheric Research, GPO Box 1538, Hobart, Tas. 7001, Australia Tel.: +61 3 62325018 E-mail: [email protected], Available under a royalty free licence, Year first available:2004. Relevant model output is available at https://figshare.com/articles/Lake_Victoria_Atlantis_model_output/3364717

Acknowledgement

Funding for this work was provided by the United Nations University - Fisheries Training Program (UNU-FTP). The authors are greatly indebted to Rebecca Gorton of CSIRO for her expertise and support.

References (40)

  • E.A. Fulton et al.

    Lessons in modelling and management of marine ecosystems: the Atlantis experience

    Fish Fish.

    (2011)
  • K. Geheb et al.

    The future of change: roles, dynamics and functions for fishing communities in the management of Lake Victoria's fisheries

    Aquatic Ecosyst. Health & Manag.

    (2007)
  • T. Goldschmidt et al.

    Cascading effects of the introduced Nile Perch on the detritivorous/phytoplanktivorous species in the sublittoral areas of Lake Victoria

    Conserv. Biol.

    (1993)
  • P. Greenwood

    The cichlid fishes of Lake Victoria, East Africa; the biology and evolution of a species flock

    (1974)
  • R.E. Hecky

    The eutrophication of Lake Victoria. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte

    Limnologie

    (1993)
  • R.E. Hecky et al.

    Multiple stressors cause rapid ecosystem change in Lake Victoria

    Freshw. Biol.

    (2010)
  • L. Kaufman

    Catastrophic change in species rich freshwater ecosystems: lessons of Lake Victoria

    BioScience

    (1992)
  • L.E.S. Kaufman et al.

    Comment the great Lakes of Africa

    Conserv. Biol.

    (2013)
  • R.B. Kelly et al.

    Selecting among five common modelling approaches for integrated environmental assessment and management

    Environ. Model. Softw.

    (2013)
  • J.F. Kitchell et al.

    The Nile perch in Lake Victoria: interactions between predation and fisheries

    Ecol. Appl.

    (1997)
  • Cited by (28)

    View all citing articles on Scopus
    View full text