SECTION A: REVISED EXPERIMENTAL DESIGNS

REPRODUCTION CONTROL

Experimental Design Modification
The funded experiment proposed to evaluate five strains of O. niloticus for variation in sex ratio. The work plan has been changed to read: The variation in sex ratio of non-hormone-treated Oreochromis niloticus stocks from a minimum of three unique origins will be described.

Experimental Design Modification
The experiment has been modified to utilize normal males and females in place of androgenetically produced males and females. Individual males will be mated with several females, and individual females mated with several males.

MARKETING AND ECONOMIC ANALYSIS

Experimental Design Modification
This work plan, as printed, included studies of economic and social returns of CRSP research in three countries (Honduras, Thailand, and the African site) over five years. The Eighth Work Plan portion examined the economic returns of CRSP research in Honduras only.

The original design of this study called for a welfare analysis approach to measure what economists term social welfare by measuring consumer and producer surplus econometrically from demand and supply estimation. However, it was not possible to gather data on a sufficiently large number of farms for a long enough time period to conduct accurate econometric analysis. An equally acceptable analytical approach to estimating returns to technology and investment is the nonparametric approach, which was instead chosen as the most appropriate given the available data. Social returns refer to the economic impacts on society of the adoption of a new technology. These returns are embedded in the results of the analysis, but with the nonparametric approach, it is not possible to separate out benefits to producers from benefits to consumers, as is done in a welfare analysis.

Experimental Design Modification
This investigation was to use cost and returns data from Thailand for fish and livestock production to construct a model of aquaculture production. The project was to model a typical farm and look at aquaculture, horticulture, and livestock alternatives, and to perform risk analyses using a Target MOTAD or a Mean-Gini model.

8MEAR2 was proposed in conjunction with 8MEAR1 as a five-year study. However, under the Eighth Work Plan, the investigation was reduced to a two-year study. It became clear during initital investigations that it would be more cost-effective to conduct 8MEAR1 and 8MEAR2 at the same site, and thus the first risk analysis was developed using data from Honduras. In Honduras, as in many other countries, there is no one typical farm and farm size often determines production and marketing strategies. Therefore, five different models were developed. Shrimp prices were used to look at effects of varying prices on the different farms modeled and on the different culture alternatives (production intensity, feeding, and fertilization alternatives) analyzed. Price sensitivity analyses were conducted to expand the research effort beyond the proposed analyses.