Exploration of flow model of aquatic ecosystems based on experimental data

Frederic Isingizwe studies how donor-control and recipient-control models fit to experimental data individually and how they compare to one another.

Introduction

It is often difficult to establish descriptions of processes or make predictions within complex systems such as ecosystem networks. It can, however, be made possible by developing models based on the evolution of flows within such systems. A network flow model is essentially an ecological food web (energy-matter flow of who eats whom). A number of models are commonly used in ecological modelling, which are mainly assumed to be linear, and they are preferred because of their simplicity. However, this linearity is still to be proven based on empirical data. This work was started with an exploratory phase, where the donor and recipient control models were studied based on experimental data. The practical objectives were to study how donor-control and recipient-control models fit to experimental data individually and how they compare to one another.

Methodology

The 7 compartments, 16 steady-state seasonal nitrogen networks from the Neuse River Estuary, North Carolina, and the St Marks National wildlife refuge (Florida, USA) Carbon Flow Networks (6 time steps, 51 compartments), were used as the experimental data for this study.

All the 51 species in the network were aggregated in seven non-overlapping non-empty subsets so that each node belongs to one and only one of them. The tests were done based on four different approaches, namely, “donor-specific donor-control model”; “recipient-specific donor-control model”; “recipient-specific recipient-control model” and “donor-specific recipient-control model”. Regression and correlation analysis were chosen as good indicators of linearity, and used in this project.

Results

Preliminary tests suggested that a network with more living compartments than abiotic ones was an appropriate choice for the study (St Marks network). Analysis yielded high correlation coefficients, in support to recipient control model in invertebrates and fish, suggesting that these recipients have higher influence on incoming flows from other compartments. Results show a bit of contradiction with the commonly, favorably used, "donor-control model." The donor - recipient control models that are usually used, are likely to be demonstrated empirically, according to the test results. In both databases (donor and recipient based) the recipient stocks have proven to be more highly correlated to the flows than it is with the donor stocks.

Conclusions

Regression and correlation analyses were conducted on empirical data of two aquatic Networks. Prior to revealing the models that are well fitted to data, an exploratory study of the state of the art was conducted, where, different combination of stocks under a power law, were regressed with respect to flows and flow ratios. Regression analysis on data from the Neuse River gave no outstanding results. Correlation analysis conducted on the St Marks network allowed a clear comparison between donor and recipient-control models, where so far, the results revealed the recipient control model to be the most reliable, as far as the extent of this study is concerned.

Supervisors

Jacek Banasiak, University of KwaZulu-Natal, School of Mathematical Sciences, South Africa
Elena Rovenskaya, Advanced Systems Analysis Program (ASA), IIASA

Note

Frederic Isingizwe of Stellenbosch University, South Africa, is a South African citizen. He was funded by IIASA's South African National Member Organization during the YSSP.

Please note these Proceedings have received limited or no review from supervisors and IIASA program directors, and the views and results expressed therein do not necessarily represent IIASA, its National Member Organizations, or other organizations supporting the work.


Print this page

Last edited: 23 March 2015

CONTACT DETAILS

Ulf Dieckmann

Principal Research Scholar Exploratory Modeling of Human-natural Systems Research Group - Advancing Systems Analysis Program

Principal Research Scholar Systemic Risk and Resilience Research Group - Advancing Systems Analysis Program

Principal Research Scholar Cooperation and Transformative Governance Research Group - Advancing Systems Analysis Program

Further information

International Institute for Applied Systems Analysis (IIASA)
Schlossplatz 1, A-2361 Laxenburg, Austria
Phone: (+43 2236) 807 0 Fax:(+43 2236) 71 313