No abstract

Residual feed intake (RFI) represents the deviation of the actual ad libitum food consumption of each animal from that predicted from combination of growth rate, backfat and metabolic body weight measurements. After 4 years of divergent selection of pigs in lines for either high or low postweaning growth rate on restricted feeding, the high line exhibited a significant reduction in RFI relative to the low line. This indicated, to some extents, a lower energy requirement for maintenance in the high than in the low line, possibly because of reduced physical activity of the animals. Estimates of genetical parameters showed that RFI was moderately heritable and those genetic correlations of RFI with carcass backfat and food conversion ratio were moderate to highly positive.

In 1990, NSW Fisheries initiated a mass selection programme in Port Stephens, NSW, with the aim of breeding faster growing Sydney rock oysters Saccostrea glomerata (Gould 1850). After two generations of selection, an average weight for age advantage of 18% (range 14±23% per breeding line) was achieved. This equates to a reduction of 3 months in the time taken to reach market size. Experiments are planned to determine how much of this 3 months advantage is additive to the 6 months advantage this laboratory has already obtained using triploid S. glomerata. A parallel set of S. glomerata breeding lines was established on the Georges River, NSW, to include selection for resistance to the protistan parasite Mikrocytos roughleyi, the causal agent of winter mortality. The programme was disrupted by the outbreak of QX disease Marteilia sydneyi, another protistan parasite, in 1994. In 1997, the breeding programme was reorganized and expanded. New lines were bred from oysters that had survived both QX and winter mortality. After one generation, a small improvement in resistance to QX has been recorded; however, the exposure of further generations to M. sydneyi will be required to con®rm an increase in resistance.

Selection was carried out in a line of pigs for increased growth rate of lean tissue. The selection criterion was weight of lean in the ham predicted from live backfat and weight measurements after a 12-week performance test commencing at 25 kg live weight. All pigs were given the same total amount of food over the test period. The scale was set to about proportionately 0-85 of predicted ad libitum intake. Boars selected with an intensity of 1/12 were used for 6 months and sows selected with an intensity of 1/4 were kept for two farrowings. An unselected control line was maintained concurrently.After five generations, performances of selected and control line pigs were compared on ad libitum and scale feeding as they grew to 85 kg. Responses in the selected line on scale feeding were +51 g/day for growth rate (GR), -0-16 for food conversion ratio (FCR), -2-2 mm for backfat (F) and +0-47 kg for ham lean (HL). On ad libitum feeding, responses were much higher in the selected line, giving rise to line x food interactions. Responses were +128 g/day for GR, -0-27 for FCR, -2-3 mm for F, +1-01 kg for HL and +0-15 kg/day for food intake (FI). Estimates of the heritability of HL from variance components were 0-43 (s.e. 0-15) on scale feeding and 0-28 (s.e. 0-19) on ad libitum feeding. The realized heritability of HL on scale feeding was 0-29 (s.e. 0-04) and its co-heritabilities with the other traits on both feeding levels were of similar magnitude to its heritability. Scale feeding exposed genetic variation in the partitioning of food between lean and fat deposition and appeared to be a suitable selection regimen for performance on ad libitum feeding. 149

This paper provides a review and summary of strain comparison and selective breeding work carried out on redclaw cray®sh Cherax quadricarinatus (von Martens) at the Queensland Department of Primary Industries, Freshwater Fisheries and Aquaculture Centre, Walkamin. Redclaw cray®sh are a highly marketable, environmentally tolerant, freshwater cray®sh with a moderately fast, but variable, growth rate. Five strains of the species were assessed morphologically and in terms of age at maturity. A non-replicated production trial was also carried out before two strains, from the Gilbert and Flinders rivers in North Queensland, were chosen to be part of an experimental selective breeding programme for improved growth. Initially, 14 families of each strain were randomly mated and grown out. At this point, a size-related selection took place, with the largest animals from each family forming a selected line, and individuals from around the mean forming a control line. Withinfamily selection and reciprocal mating between families was performed. Data were collected after each of two generations of growout and analysed by analysis of variance. Signi®cant differences (P < 0.05) were measured between strains, sexes and between selected and control lines. Selected individuals grew 9.5% faster than the controls. Recommendations for selective breeding research involving freshwater cray®sh are presented.