Modeling of aquaculture pond systems has been a continuing activity of the PD/A CRSP. Over the years models have been developed to simulate water quality parameters under a variety of conditions. Initial models were built under the assumption that pond conditions were approximately uniform. Later on, models were developed such that stratification events could be simulated over short periods of time. Those models were further refined so that long-term simulations of up to a full growing season could be carried out. In doing these long-term simulations, the models were enhanced to account for stochastic variations in weather. Other models have been developed to look at nutrient cycling in integrated aquaculture–agriculture systems. All the models developed to date at UC Davis have been implemented using STELLA™, a commercially available modeling language available for both Macintosh and Windows operating systems. Although the use of STELLA™ for model development greatly simplified and accelerated the process of model programming, it has certain limitations for model distribution and use by interested individuals. Under the work proposed for the Ninth Work Plan, a PD/A CRSP model previously developed at UC Davis will be updated and re-coded to create a “user-friendly” version that can be run without specialized programming knowledge or software packages.

Collaborating Institution
University of California, Davis

Raul H. Piedrahita
Zhimin Lu

Introduction
The work being proposed here expands on models previously developed at UC Davis. The objective in this work plan is to enhance the usefulness of previous models by improving the accessibility and realism of pond water quality, fish yield, and effluent simulations. Nutrient, especially nitrogen and organic matter, dynamics will be considered in a model which also integrates stochastic features. In a shift from previous models developed at UC Davis, the new model will be specifically developed with a user-friendly interface, and in a form that can be easily distributed to users.

The work being proposed under this work plan complements work by the OSU Biosystems Analysis Group. The model to be developed at UC Davis will have user interfaces developed in consultation with the OSU Biosystems Analysis Group in an attempt to make those interfaces compatible with the POND^© interfaces. The UC Davis model will emphasize characteristics such as stochastic generation of weather parameters, stratification, and detailed nutrient cycles. The UC Davis model will also have as an objective the evaluation of potential environmental impacts by analyzing water and nutrient releases as affected by environmental conditions and by pond management practices.

Objectives
The specific objectives of the proposed work are in accordance with the objective listed in the Continuation Plan for the Aquaculture Systems Modeling Research Theme: “To analyze and synthesize research results into models which better describe system processes.” The specific objectives are:

1) To develop a stochastic model for estimating water quality and fish growth in fish ponds as determined by climate, pond management practices (feeding and fertilization), and quantity and quality of the water supply. The model will be of use in obtaining probability distributions for estimates of water quality, fish yield, and nutrient and water releases from ponds.
2) To program the model in a form that is “user-friendly” and can be run without specialized programming knowledge, or software packages.

In addition to supporting the Aquaculture Systems Modeling Research Theme, the objectives also support the Effluents and the Appropriate Technology Research Themes.

Significance
The development of models to predict fish yields and effluent water quality from aquaculture operations is an important tool in the process of improving the sustainability of aquaculture (Briggs and Funge-Smith, 1994). The models can be run with climatic and other site-specific data in addition to pond management information, to determine the potential aquaculture yields, and nutrient and water discharges. Stated objectives of the Continuation Plan include the reduction of effluents and the development of appropriate technologies, and the use of a model such as the one being proposed here can result in progress towards those objectives.

The potential impact that aquaculture may have on the environment has received considerable interest recently by scientists, regulators, and aquaculturists (Briggs and Funge-Smith, 1994; Schwartz and Boyd, 1994; Kelly et al., 1994; Hopkins et al., 1993). Furthermore, previously aquaculture had an image as an environmentally benign practice but this has changed in recent times. At the 1997 meeting of the World Aquaculture Society in Seattle, there were demonstrations and “counter conferences” organized by a variety of activist environmental groups. The increased interest in the environmental impacts of aquaculture has been due, in large measure, to the rapid growth of aquaculture and the environmental degradation that has occurred in mangrove areas and in estuaries through shrimp farming and other aquaculture practices (Pillay, 1992). Effluent discharge from aquaculture may have deleterious effects on the biota in receiving waters through oxygen depletion, increased suspended solids concentration, temperature changes, diseases, and eutrophication (Ackefors and Enell, 1994). Application of aquaculture effluents to agricultural land can be used to enhance the growth of crops. Land application of aquaculture effluents can also reduce the environmental impact of discharged effluents by removing suspended solids and nutrients. Salt concentrations in some aquaculture waters and scheduling problems in matching the irrigation and nutrient needs of the terrestrial crop with the availability of aquaculture effluents are some of the difficulties encountered in using aquaculture effluents in agriculture.

Modeling work being conducted at UC Davis under the Eighth Work Plan has shown that organic matter and nitrogen dynamics in the sediment and water column of waste-fed aquaculture ponds can be reliably predicted using mechanistic models. In other work at UC Davis, it has been shown that temperature, dissolved oxygen, and fish growth can be predicted using stochastic methods. The work being proposed for the Ninth Work Plan will involve the linking of the stochastic temperature and DO model with the water column and sediment organic matter/nitrogen dynamics model. However, previous models developed by the UC Davis DAST have been programmed using STELLA™, a specialized modeling language. Whereas the use of STELLA™ has greatly facilitated the model development process, the resulting models are not in a form that can be easily distributed and used by non-modelers. Therefore, the combined and enhanced stochastic water quality and fish yield model will be coded using a programming language such as C++ so that user-friendly interfaces can be incorporated, and the model can be distributed for execution without the need for specialized software.

Anticipated Benefits
Much of the modeling work carried out at UC Davis has received relatively little exposure and use outside the realm of researchers and modelers. In this work plan, we intend to integrate our past efforts in the development of a stochastic model, and of the detailed nutrient cycling included in the pond model that is part of the coupled aquaculture/agriculture systems model. Furthermore, the final version of the model developed under the Ninth Work Plan will summarize the mechanistic models of aquaculture ponds developed at UC Davis in a user-friendly and accessible tool. The new modeling tool will be designed to be used to assess carrying capacity, potential fish yields, and effluent flow rates and characteristics at a given site. The target users for the new model will be primarily researchers, designers and aquacultural development planners.

Activity Plan
The work to be carried out under the Ninth Work Plan has two major components. One component is the integration and enhancement of models previously developed at UC Davis. The second component is the re-formulation of the combined model into a form that is easily distributed and used by non-modelers. Details of the research plan are presented below.

Location of Work: All programming work will be carried out at UC Davis. A one week annual working visit to Corvallis will take place. The meeting will be used to coordinate the work with the OSU Biosystems Analysis Group.

Research Plan and Methodology: During the Eighth Work Plan, and under the general heading of Aquaculture Pond Modeling for the Analysis of Environmental Impacts and Integration with Agriculture, the UC Davis group has been working on two studies: ASMR1) Relationship Between Carbon Input and Sediment Quality in Aquaculture Ponds, and ASMR2) Stochastic Modeling of Temperature, Dissolved Oxygen and Fish Growth Rate in Aquaculture Ponds. The work to be undertaken in the Ninth Work Plan will have two distinct components related to the specific objectives listed above:

1) To develop a stochastic model for estimating water quality and fish growth in fish ponds as determined by climate, pond management practices (feeding and fertilization), and quantity and quality of the water supply. The two models developed under the Eighth Work Plan will be combined into a single model. In the first Study under the Eighth Work Plan, a model has been developed to simulate the flow of nitrogen and of organic matter between an aquaculture pond and a coupled terrestrial crop system. Although the model includes sediment/water interactions, this is an area for which there is very limited information available. Field work currently in progress under the PD/A CRSP Eighth Work Plan will provide information which will complement and enhance previous knowledge on pond sediments and how they interact with the water column. This information will be used to update the simulation of sediment/water column interactions.

The new, integrated model will have the detailed water column/sediment interactions and nutrient relationships included in the Eighth Work Plan Study ASMR1 model, as well as the stochastic weather generation and solution procedures developed for the Eighth Work Plan, Study ASMR2 model. The new model will also account for pond stratification, as in the Eighth Work Plan, Study 2 model. Initial model integration will be carried out in STELLA™, the simulation modeling language used for development of the Eighth Work Plan models.

2) To program the model in a form that is “user-friendly” and can be run without specialized programming knowledge, or software packages. The OSU Biosystems Analysis Group has been very successful with their development of POND^©. They now have had over 900 downloads of the program from their Internet web site (J. Bolte, personal communication, September 1997). Their success is due in large measure to their efforts at making their work accessible by creating user-friendly interfaces for their models. We will work with the OSU Biosystems Analysis Group in developing the architecture for the new model coding, and in endeavoring to make the user interfaces consistent with the POND^© interfaces. The annual meeting with the OSU Biosystems Analysis Group will be used to explore issues related to model architecture, to determine the extent to which some of the code developed for POND^© (for actions such as data input and output, file creation and management, etc.) can be used in the new model, and to maintain consistency in user interfaces to take advantage of the POND^© user base.

Regional Integration
The proposed research is “cross-cutting” and will apply to all the regions in which the PD/A CRSP has been active. Data used for model development, calibration, and validation will be from the PD/A CRSP Database. The model will be tested with data from all the regions.

Relative Contribution to Geographical Regions
The models developed will be most useful in regions for which data are available for model calibration and validation. Given the extent of the PD/A CRSP involvement, and the magnitude and scope of the Data Base for the various regions, the relative contribution is estimated as: 30% SE Asia, 30% Africa, 30% Central America, and 10% South America.

Schedule
Integration of the models will take place during the first year, and a complete, calibrated version will be available by April 2000. This initial version will be developed in STELLA™. A parallel effort at programming the models in C++ will be started in May 1999. By April 2000 a preliminary version of the integrated model and a user’s manual will be available for distribution and testing. Comments and suggestions received from users will be incorporated in a version for general distribution by April 2001.

Report Submission
Progress reports will be submitted annually. A user’s manual for the model will be available for distribution with a preliminary version of the software by April 2000. A revised version of the software and of the manual will be available for distribution by April 2001.

References
Ackefors, H. and M. Enell, 1994. The release of nutrients and organic matter from aquaculture systems in Nordic countries. J. Appl. Ichthyol., 10:225-241.
Briggs, M.R.P. and S.J. Funge-Smith, 1994. A nutrient budget of some intensive marine shrimp ponds in Thailand. Aquacult. Fish. Manage., 25:789-811.
Hopkins, J.S., R.D. Hamilton II, P.A. Sandifer, C.L. Browdy and A.D. Stokes, 1993. Effect of water exchange rate on production, water quality, effluent characteristics and nitrogen budget of intensive shrimp ponds. J. World Aquacult. Soc., 24:304-320.
Kelly, L.A., A. Bergheim, and M.M. Hennessy, 1994. Predicting the output of ammonium from fish farms. Water Res., 28:1403-1409.
Pillay, T.V.R., 1992. Aquaculture and the Environment. Fishing News Books, London. 189 pp.
Schwartz, M.F. and C.E. Boyd, 1994. Channel catfish pond effluents. Prog. Fish-Cult., 56:273-281.