| Sidenavigasjon: Hovedside / Institutter / Husdyr- og akvakulturvitenskap |
[Sidekart] [Kontakt] 
|
| Tekststørrelse. |
Thomas Gitterle - doktorgradsavhandling
Ane Gro Siri SkjelfjordAbstract fra avhandlingen
Genetic analyses for resistance to White Spot Syndrome Virus (WSSV), harvest body weight and pond survival in the Pacific white shrimp Penaeus (Litopenaeus) vannamei
English language abstract
The shrimp farming industry is still mainly supplied with larvae produced from broodstock captured from the wild, which represents a serious weakness for the industry. Firstly, because the use of wild stocks poses a serious risk of introduction of disease and secondly because the advantages of domestication and genetic improvement cannot be utilized.
In this thesis, the establishment of a commercial breeding programme for Penaeus vannamei is described. This included the production and rearing of full- and paternal half-sib families and their evaluation for harvest weight and pond survival under both semi-intensive commercial and intensive tank rearing conditions. The families were also tested for resistance to White Spot Syndrome Virus in experimental challenge tests. The reports of the genetic (co)variances estimates for these three traits are presented for the first time.
Heritability estimates for harvest weight ranged from 0.01±0.09 to 0.54±0.022 and for pond/tank survival from 0.02±0.02 to 0.12±0.06. Results also indicate a positive genetic correlation between these two traits and a low genotype by farm interaction. In the first attempt to estimate the genetic variances for resistance to WSSV we applied a linear animal model that defines resistance as a binary trait (dead or alive) based on whether an animal is still alive or not when the population reached 50% mortality. Data were obtained by experimentally infecting the animals by voluntary feeding with minced muscle tissue from shrimps that died of WSSV infection. The heritability estimates for this trait were low (range from 0.03±0.02 to 0.017±0.06). Further more we also found a negative genetic correlation between this trait and harvest weight (from -0.16±0.61 to -0.94±0.64). It was however, hypothesized that both the applied statistical model for the genetic estimations and the challenge test protocol could have affected the magnitude of the heritability estimates for resistance to WSSV and also the genetic correlation between this trait and harvest weight. The model did not include time from infection to death of the animal and the challenge test protocol used not ensures that all animals were equally at risk of being infected at the same time. The unfavourable genetic correlation between harvest weight and resistance to WSSV could be a consequence of the challenge test protocol where bigger animals tend to be more voracious and hence get infected first and with a higher viral load. Therefore more sophisticated statistical models and alternative challenge test protocols were evaluated.
Five statistical models were evaluated for the estimation of variance components and prediction of breeding values. With a Cox proportional hazard and a linear repeatability model selection accuracies were higher compared with the initial linear model. However the heritability estimates were low with all five models. To optimize the challenge test protocols two alternative routes of infection, individual oral (IO) and waterborne (WB) infection were evaluated. The genetic variance were estimated using the above mentioned linear repeatability model. With the IO protocol it was possible to have a better infection control as well as equaling the risk of infection in all animals and at the same time. With the second property the time period since the animals are exposed to the virus and the event time (failure or censoring) is known. This is particularly important for a proper use of the statistical models that measure disease resistance using time to death as dependent variable like the Cox proportional hazard and the linear repeatability model.
The genetic correlation between WSSV resistance and harvest weight remained unfavorable in most of the batches when animals were IO infected. The IO infection protocol should avoid the effect of animal behavior on the estimated genetic correlation.
Pr. mai 2005:
Thomas Gitterle, Acuanal / Ceniacua, Cra 8A no 96-60, Bogotá, Colombia
Publisert: 01.12.08
Oppdatert: 06.01.09
Utskriftsvennlig versjon
Del med en venn:
Institutt for husdyr- og akvakulturvitenskap