III. Global Research

The CRSP conducted global research on the following research themes: Pond Dynamics, Reproduction Control, Marketing and Economic Analysis, Adoption/Diffusion, Aquaculture Systems Modeling, and Decision Support Systems. Auburn University PD/A CRSP scientists elaborated on a classification system to describe pond soil horizons. Soil samples collected from CRSP sites in Thailand and Honduras were analyzed and compared for moisture content, dry bulk density, color, wet and dry soil pH, exchangeable acidity, particle density, total carbon and nitrogen, total phosphorus, total sulfur, and acid-extractable phosphorus and metal ions. Results of soil analyses verified that the classification system could be used to delineate pond soil horizons at a diversity of sites. This system, which provides a framework for the characterization of pond soils, promises
to yield soil descriptions that are essential for future PD/A CRSP research and will assist in the refinement of pond management strategies.

Currently PD/A CRSP reproduction control research is aimed at developing monosex tilapia populations through the investigation of immersion treatments for the masculinization of tilapia and androgenesis techniques for the production of YY male tilapia. Several advantages are associated with the monosex production of tilapia: fish growth is enhanced, unwanted tilapia reproduction is prevented, and yields are increased. One way to masculinize tilapia is to administer hormones in their diet; however, the following problems can occur with the use of this technique: 1) uneven hormone exposure of tilapia; 2) high variability in the percent males produced; 3) worker exposure to steroids (via contact with skin or through the air while feeding); 4) steroidal contamination of the pondwater and sediment and/or the surrounding terrestrial environment; 5) lengthened period of time required for masculinization; and 6) steroid exposure of untargeted organisms. An alternative sex reversal technique that may address the limitations of hormone-treated feed is the immersion of tilapia fry in steroid solutions. Previous CRSP studies of steroid immersion demonstrated that immersion for three hours in 17a-methyldihydrotestosterone (MDHT) on days 10 and 13 post fertilization resulted in the masculinization of Nile tilapia.

PD/A CRSP studies continued to focus on short-term immersion therapy at Oregon State University (OSU) and progressed according to schedule. Researchers tested the effects of: 1) rearing density on the efficacy of MDHT immersion; 2) MDHT immersion treatments applied on a single day post fertilization rather than on two different days post fertilization; and 3) immersion treatments using trenbolone acetate (TBA). Results of this study supported outcomes of previous studies that successfully demonstrated the use of short-term immersion treatments and revealed that a single immersion in MDHT on day 13 post fertilization was an effective procedure for masculinizing tilapia. Findings of this study in comparison with previous research results also suggest that masculinization outcomes may improve with longer duration immersion treatments. Tilapia were stocked at 33, 67, 100, and 200 fish l^-1 and immersed in 500 µg l^-1 MDHT. Stocking density also affected the sex reversal outcomesa stocking density of 33 fish l^-1 resulted in 80.3% males, whereas the 67 and 100 fish l^-1 stocking densities resulted in 71.1% males, and the 200 fish l^-1 stocking density resulted in 64.5% males. Two 2-hour immersions in TBA resulted in greater than 90% males in a female-biased brood as compared with 20% males produced in two 2-hour immersions in 17a-methyltestosterone (MT). Additionally, experimental results suggest that the period of sensitivity to steroid-induced masculinization for Oreochromis niloticus occurs several days after the initiation of feeding.

Scientists at the University of Oklahoma submitted a progress report describing work geared toward the development of androgenesis techniques applicable to tilapia. Research was delayed and will overlap with the period of the Ninth Work Plan and only initial experimental
results are presented regarding the production of androgenotes. To ensure that freshly produced gametes are available for chromosome manipulation, the effects of photoperiod manipulation on the spawning of Nile tilapia (Oreochromis niloticus) were evaluated. Tilapia ovulated 9.5 to 13.5 hours after a controlled, northwest, light-on cycle of a 20 L:4 D photoperiod, which confirmed that tilapia ovulation is light cycle sensitive and that photoperiod manipulation is a viable technique for determining stripping time. To perform chromosome manipulations, researchers also developed a tau curve based on a temperature-related rate of development for tilapia. A reasonable first estimate of a tau curve was derived for Nile tilapiathe mean mitotic interval during early synchronous cleavage was between 73.5 and 30 min at 20.6 and 27.5°C, respectively. Future research efforts will be directed toward refining androgenesis methodology and will include identifying more suitable egg incubation techniques, standardizing the thermal shocking of eggs using the working tau curve, determining UV dosage for the deactivation of the female genome, and selecting broodstock with a visual genetic marker for future androgenetic studies.

Marketing and Economic Analysis research activities at the University of Arkansas, Pine Bluff, are progressing according to schedule. PD/A CRSP scientists are presently involved with a study focused on technology adoption and the quantification of economic and social returns associated with CRSP-developed technologies. Many factorschanging market structures, the cost and returns of a given technology, environmental conditions, and established farm practicesinfluence a fish farmer's decision to adopt a new technology. Researchers are seeking to improve the integration of CRSP technologies and are developing a "safety first" model that will analyze, through the use of survey data collected from shrimp and tilapia producers, the integration of pond fertilization schemes into shrimp and tilapia farming systems in Honduras. The model addresses the question of the sustainability of fertilization schemes by encompassing a broad range of perspectivesthe environment, economics, known risk factors, and food security.

To quantify the social and economic returns associated with the adoption of CRSP technologies by Honduran shrimp and tilapia farmers, researchers have designed impact and welfare analysis models to evaluate two specific CRSP technologies: feed regimes recommended to shrimp producers and sex reversal techniques utilized by tilapia producers. The results obtained from these welfare economics models are potentially useful in developing policy recommendations and in evaluating the social impact of CRSP research.

Pond aquaculture models can be used to compare the effects of different management regimes on fish yields, the economic efficiency of a production system, and the efficiency of resource use. Data Analysis and Synthesis Team (DAST) researchers at OSU continued according to schedule with two studiesone to evaluate and compare two areas of production technology and a second to further refine POND^© software. The first study compared fertilization strategies (fixed input versus responsive strategies that were generated by PONDCLASS^© pond management software) and feed application rates introduced in Thailand, the Philippines, and Honduras. PONDCLASS^© fertilization rates were three to seven times more efficient in terms of phosphorus recovery in fish flesh. The efficiency of nitrogen recovery for responsive strategies was comparable to or slightly higher than fixed input strategies. Fixed input fertilization strategies in general resulted in higher net fish yields; however, treatments receiving responsive fertilization rates were one and a half to three times more cost efficient than fixed input fertilization rates in the Philippines and Thailand and comparable to fixed input rates in Honduras. Results of the comparative analysis showed that fixed input fertilization rates, specifically phosphorus inputs, at CRSP sites could be reduced without compromising production efficiency. To develop feeding schedules that minimize feed use, an adaptive non-linear search strategy implemented in POND^© software was used to generate optimal feeding schedules requiring less feed for individual ponds. Simulation results indicated that with minimized feed use culture periods tended to be longer in comparison with satiation feeding. Whether minimum feed use results in net savings requires further analysis; the increased costs of longer culture periods may offset reduced feed costs.

The second research effort of the OSU DAST was directed toward the continued refinement of POND^© software and the assessment of PD/A CRSP production technology. To improve the user interface of POND^© software and provide increased decision-making support, computer automation tools were implemented to automate processes
such as setting up ponds and lots, generating liming and fertilization guidelines, developing optimal feeding schedules, and simplifying the steps required to perform simulations under varied conditions. POND^© software models were also upgraded to account for the effects of: 1) fertilization and high fish biomass on fish growth; and 2) the effects of different feed types in terms of moisture, protein, and energy content on fish performance. Additionally, DAST researchers are designing models for POND^© software that will simulate phosphorus flux in pond water and sediments and polyculture interactions in ponds.

DAST research at the University of California, Davis (UCD) is progressing according to schedule with the continued improvement of an integrated aquaculture/agriculture model developed from an existing aquaculture pond ecosystem model and modification to several components of a model designed to predict fish growth in stratified fish ponds using stochastic weather variables. Improvements incorporated into the integrated aquaculture/agriculture model included the explicit consideration of organic matter and nitrogen transformations, the incorporation of the effects of low quality feed on fish growth in the fish growth model, the inclusion of sediments in the mass balance equations, and the coupling of an agricultural component to the aquaculture pond ecosystem model. The most recent developments involved incorporating sediment mineral processes, calculating the carbon to nitrogen ratio in the water column, and modeling the light extinction coefficient in ponds with high non-algal turbidity. Data relating to environmental conditions, feed and fertilization rates, and stocking densities from the Butare Rwanda CRSP site were used to run the model, and the three ponds used for model calibration received a weekly input combination of green grass, urea, and chicken manure. Researchers tested the improved accuracy of parameters relating to water column organic matter, water column total ammonia nitrogen, sediment organic matter and sediment total nitrogen, and chlorophyll a. Simulated and observed values for chlorophyll a, total ammonia nitrogen concentrations in the water column, and sediment organic matter and total nitrogen were comparable. Results of this experiment also indicated that an increased level of computational detail in the water column and sediment processes is necessary to improve the accuracy of simulation of organic matter and nitrogen.

In addition to improving the integrated aquaculture/agriculture model, DAST researchers at UCD further modified several components of a model designed to predict the effects of random weather variables on water temperature, dissolved oxygen, and fish growth in stratified fish ponds. The model is now better able to quantify phytoplankton respiration. It can factor in pond water exchange and assess variables such as different types of organic matter with varying rates of decomposition and oxygen consumption. The fish growth model was also modified to allow for the consumption of different foods. Data from the PD/A CRSP Rwanda site were used to test the model, which was run 20 times for a simulation period of 146 days. Results of simulations indicated a high degree of stratification in the water column of ponds, which may have been due to a high light extinction coefficient and high phytoplankton concentrations. Simulated average fish weight and observed fish weight were in close
agreement. Results also showed that the probable range of expected fish size increased as the duration of the growing period increased.

Since its beginning the PD/A CRSP has provided either direct or indirect funding for students seeking undergraduate and graduate level degrees related to the field of aquaculture. Adoption/Diffusion research regarding the human capital impacts attributable to the Global Experiment is delayed. Scientists from Auburn University are initiating an exploratory study to determine: 1) students' motivations for pursuing degrees related to aquaculture; 2) the career path of students since they completed their degrees; 3) the extent of aquaculture information disseminated in students' communities as a result of their education; and 4) the students' contributions to CRSP project goals during their academic careers and after graduation.