Abstract
Taiwan Red-feathered country chickens (TRFCC) was the main livestock products in our country. Owing to the lack of any systematic breeding program, the egg production rate has gradually decreased. Based on the basis of proteomics, this study was to investigate the correlation between serum protein levels and egg productivity during the laying period. Furthermore, multiple-stage selection for egg production using serum protein levels was proposed from growing to the laying period. The thesis was segmented into three chapter：chapter 1, analysis of oviposition-related biopathway in avians; chapter 2, association of serum protein levels with egg productivity in Taiwan Red-feathered country chickens; chapter 3, selection strategy using serum protein levels to improve egg production in Taiwan red-feathered country chickens. Three groups of serum proteins, including sex hormones and cytokines, egg yolk-related molecules and egg white-related molecules were related to egg production. In order to understand the association between serum protein levels and egg number, we investigate the relative levels of four ovipostion-related proteins, vitellogenin, ovotransferrin, IGF-I and apo A-I during initial egg production (24 wk) and peak egg production (35 wk). Except for ovotransferrin, the other proteins were shown to be significantly associated with egg number at peak egg production compared to the number of initial egg production. Since serum protein levels were not significantly different at 24 wk old, traditional statistical analysis using serum protein levels as a selection threshold could not be applied for chicken selection at this stage. However, selection at latter stage using serum protein levels seems not to decrease the breeding costs. To solve the problem, we bring up code-selection method by observing the levels of serum proteins in two batches of birds at 14, 24 and 35 wk old. The approach was to transform serum protein levels as a set of scores and ranks. The set of scores and ranks was then combined as a set of regional codes. Then the birds with similar codes in two batches of birds were selected. By the code-selection approach, we have found the regularity of serum protein levels of low egg production populations in these two batches of birds. 19.5% of chickens could be selected and 3% of egg production rate was estimated to be improved by the method during 14 wk old. Moreover, within those selected birds, 78.8% were belonged to lower 50% egg production group. The novel selection strategy of those protein levels at immature stage may provide a screening model to decrease the breeding costs and improve egg productivity.