Muhammad Hunaina Fariduddin Aththar

Determining the breeding goal is the first and important step for designing a breeding program. We calculated the economic value of LnVar to determine its importance in breeding goals. LnVar was calculated by fitting the expected individual growth curve based on longitudinal observed weight. In aquaculture practices, feed requirements are predicted based on expected fish weight estimated from periodic sampling of groups of fish. It can be hypothesized that deviations of actual weight from expected weight will lead to economic losses. In Chapter 3, we derive the economic value of LnVar and explore the potential of economic gain from reducing LnVar using selective breeding. To calculate the economic value of LnVar, we define the effect of fluctuations in fish growth as the economic loss resulting from feed waste, growth deficiency and feed saving. The resulting economic value (EV) for LnVar is 0.043 US$/unit LnVar/kg production. The breeding program to improve HW and LnVar with the selection index only on HW showed a total economic response of 0.110 US$/kg per generation, whereas incorporating LnVar into the index alongside HW increased the response to 0.122 US$/kg, showing approximately 11% improvement in economic response. Further, to investigate the economic response of including LnVar alongside HW in a breeding program to improve both traits, we simulated two scenarios: one with selection index including only HW, and another with index that included HW and LnVar. Incorporating LnVar into the index alongside HW in the breeding program to improve HW and LnVar increased the economic response by more than 11% compared to the selection index including only HW. Therefore, we recommend that fish breeding programs collect repeated records of body weight and include LnVar in the breeding goal.

In Indonesia, recurrent farming failures due to disease outbreaks have driven shrimp farmers to develop co-culture between shrimp and tilapia. Shrimp farming takes place largely in brackish water areas, such as the north coast of Java. To produce tilapia in this area, farmers need a fish that can be exposed to salinity fluctuations experienced in their ponds. The Research Institute for Fish Breeding (RIFB) Indonesia has been conducting a small-scale breeding program for salinity tolerance using the Sukamandi tilapia to develop a fast-growing tilapia with good growth over a range of fluctuating salinities in brackish water ponds. A freshwater nucleus and evaluation breeding program is the simplest strategy to implement but requires knowledge of the extent of the genotype-by-environment (GxE) interaction between fresh and brackish water environments. In Chapter 4, we investigate the impact of salinity on genetic parameters and the presence of GxE between brackish water and freshwater ponds in Sukamandi tilapia. The results showed that brackish water ponds provide better support for fish growth, resulting in higher growth performance. We conclude that there is substantial GxE interaction for growth between brackish water and freshwater. We recommend that a breeding program for salinity-tolerant tilapia with a safe, stable, low-risk, and bio-secure freshwater nucleus should incorporate sib information on growth performance in brackish water.

Furthermore, to increase production in brackish water environments, fish farmers need resilient tilapia capable of consistent and predictable growth performance. In Chapter 5, we estimated genetic parameters for growth and for LnVar in the Sukamandi tilapia in brackish water. We produced 102 tilapia families and randomly assigned fingerlings to grow-out in co-culture with shrimps or to grow-out in monoculture. Results showed that Sukamandi tilapia is able to thrive in co-culture with shrimps within brackish water environments, achieving growth rates comparable to monoculture. We found heritable variation in LnVar for tilapia grown in the brackish water ponds. We found moderate GxE between co-culture and monoculture. This suggests that genetic variation for growth consistency is expressed in the presence of shrimp. The magnitude of GxE for LnVar between co-culture and monoculture is higher than that for the growth parameters, suggesting that LnVar is more responsive to the environmental differences than growth. The genetic correlation between LnVar and the growth of Sukamandi tilapia is less than unity, which supports the idea that LnVar and growth are different traits. To enhance predictable fish growth in the brackish water environment, we recommend that fish breeding programs collect repeated records on body weight and include both LnVar and growth in the breeding goal, assigning appropriate weights to each trait.

In Chapter 6, I present a broader discussion on the concept of growth consistency as an indicator for resilience and its application in fish breeding programs. I start by describing the sources of stressors in aquaculture, followed by a discussion on growth consistency, resilience and environmental sensitivity.