dans lesquelles les rétrotransposons jouent un rôle moteur. Dans ce cadre, nous nous sommes fixé trois objectifs de travail : 1) améliorer notre connaissance des relations phylogénétiques au sein du genre Lupinus (Fabaceae) par l'utilisation de nouveaux marqueurs nucléaires (ARNr-ETS et SymRK), 2) évaluer par amplification et par hybridation in situ la diversité, l'abondance et le rôle des rétrotransposons Ty1/copia et Ty3/gypsy dans les variations de taille de génome des lupins, et 3) séquencer, annoter et comparer une première région génomique disponible pour un lupin avec les régions homologues d'autres fabacées. La phylogénie obtenu améliore notre compréhension de l'histoire évolutive des lupins, etmet en évidence des schémas de variation de taille de génome différents d'une lignée à l'autre. Les analyses de rétrotransposons révèlent que les éléments copia et gypsy contribuent de façon plus significative aux différences de taille de génome chez les lupins méditerranéens que chez les lupins africains et suggèrent différents modes et mécanismes d'évolution de la taille des génomes au sein du genre. À l'échelle locale (région du gène SymRK), nous confirmons la forte implication de ces éléments qui représentent 25% de la région analysée chez Lupinus angustifolius.

The nutritional value of faba beans (Vicia faba), Australian sweet lupins (Lupinus angustifolius), white lupins (Lupinus albus) and peas (Pisum sativum) grown in New Zealand for broilers were evaluated in terms of their nutritional characteristics, protein quality (protein efficiency ratio), apparent metabolisable energy, apparent ileal digestibility coefficient of amino acids and the effects on bird performance. The effects of dehulling and extrusion cooking on the nutritive value of legumes were also investigated. The first experiment discussed in Chapter 3 evaluated the effect of cultivars on the nutrient profile and protein quality of chickpeas (Cicer arietinum), Australian sweet lupins, peas and soybeans (Glycine max). With the exception of white lupins, cultivars had no effect on the proximate and fibre composition of grain legumes. Starch was the primary carbohydrate component of chickpeas and peas, whilst non-starch polysaccharides were the major carbohydrates in lupins. The legume proteins were deficient in lysine, methionine, cystine and threonine. No differences were found in protein quality between cultivars of the different grain legume species. The lowest weight gain and protein efficiency ratio, in addition to the highest relative pancreatic weight and mortality rate was found in raw soybeans, suggesting that soybeans contained high a concentration of anti-nutritional factors, such as protease inhibitors. Birds fed chickpeas, lupins and peas had a low mortality rate and relative pancreatic weight, confirming that the level of anti-nutrients in these legume seeds was low. The apparent metabolisable energy and apparent ileal digestibility coefficient of amino acids of faba beans, Australian sweet lupins, white lupins and peas were determined in the second experiment (Chapter 4). Cultivar effect on the apparent metabolisable energy values was observed only for faba beans and white lupins. Faba beans, white lupins and peas had comparable apparent metabolisable energy values, but these values were higher than those of Australian sweet lupins, and lower than that of soybean meal. No cultivar differences were found in the apparent ileal digestibility coefficient of amino acids of grain legumes. The apparent ileal digestibility coefficient of amino acids of both lupin species was found to be comparable to that of soybean meal. The effects of feeding diets containing 200 g/kg faba beans, lupins or peas on the performance, digestive tract development and litter quality of broilers were investigated in the third and fourth trials. In the cage trial (Chapter 5), the results showed that the weight gain of birds fed diets containing grain legumes was similar to that of control diet. Feed intake and feed per gain of birds fed diets containing the majority of grain legume cultivars did not differ from those fed the maize-soy diet. Birds fed diets containing faba beans had more dry and friable excreta compared to other treatment diets. The performance of birds fed diets containing 200 g/kg grain legumes during the 35 d grow-out period, in the floor pen trial (Chapter 6), confirmed the results of the cage trial. In this trial, weight gain and feed per gain of birds fed diets without meat meal were superior to those with meat meal. In cage trials, the modification of some segments of digestive tract development was probably due to the dietary NSP. Whilst in floor pen trial, digestive tract development was not influenced by the inclusion of grain legumes. The effect of methodology of determination (direct vs. difference method) on the apparent ileal digestibility coefficient of amino acids of wheat, maize, Australian sweet lupins, peas and soybean meal for broilers was evaluated in the fifth study (Chapter 7). The influence of methodology on apparent ileal digestibility coefficient of amino acids was found to vary amongst the feed ingredients. In general, the apparent ileal digestibility coefficient of amino acids of test ingredients determined by the difference method was higher than those determined by the direct method, suggesting that the use of the direct method may underestimate the apparent ileal digestibility coefficient of amino acids in low and medium protein ingredients. Data reported in Chapter 8 shows that dehulling increased the apparent metabolisable energy values of faba beans and Australian sweet lupins, but it had no beneficial effect on peas. The increase of apparent metabolisable energy values may be attributed to the decrease in non-starch polysaccharides of these legume seeds after dehulling. The removal of hulls increased the amino acid concentrations, but it had no effect on the apparent ileal digestibility coefficient of most amino acids. These results suggest that dehulling of grain legumes would be nutritionally beneficial and, likely to be economical in view of the improved amino acid concentrations and energy values. The final experiment (Chapter 9) demonstrated that extrusion of peas markedly influenced the content of crude protein, non-starch polysaccharides, starch, and trypsin inhibitors. The soluble non-starch polysaccharides and trypsin inhibitor contents of the majority of extruded pea samples were higher than those of raw peas, but insoluble and total non-starch polysaccharides decreased with extrusion. Extrusion had no effect on the apparent ileal protein digestibility and the apparent metabolisable energy of peas, but it increased ileal starch digestibility.

[Truncated abstract] Native Australian soils contain very low amounts of phosphorus. The soils of southwestern Australia are ancient and highly weathered. Consequently, the availability of phosphorus in these soils is too low for cropping purposes, so the application of P is necessary to maintain productivity. When P is applied to soil, typically as soluble superphosphate, it tends to be transformed to increasingly less soluble forms over time. Sparingly soluble forms of soil P are relatively inaccessible to Triticum aestivum; however, many grain legumes have a higher P-acquisition efficiency, allowing them to access pools of soil P that T. aestivum cannot. The P-acquisition efficiency of some grain legumes has been attributed in part to their ability to release large quantities of carboxylates, coupled with the development of cluster roots for species such as Lupinus albus. There are a number of unexplained observations in terms of the P-acquisition efficiency of grain legume species and the way that those species respond to P fertilisation. This PhD project aimed to study carboxylate release from a range of crop species, and investigate its role in variation among species for acquisition of phosphorus from sparingly soluble forms (chapter 1). ... L. albus (chapter 5). There was considerable variation in P acquisition among accessions. The variation cannot be attributed to differences in carboxylate release, cluster-root development or whole root system rhizosphere extract pH as measured in this study. We hypothesise that the variation might be attributed to differences in the ratio of release of protons and other cations localised around cluster roots. In conclusion studies of carboxylate exudation and sparingly soluble forms should use more than a single form if the aim is to draw generalised conclusions on P-uptake efficiency from sparingly soluble forms. Comparative studies of a range of species are a useful tool for enhancing our understanding of root physiology. While the benefit of carboxylates for providing access to poorly soluble P has been demonstrated, questions remain as to potential other roles for carboxylates, particularly in species that do not form cluster roots. Variation in P uptake among accessions of L. albus is present, and more work on proton release and ion balance of root clusters is necessary to understand intraspecific variation.

Two trials using Lupinus angustifolius cv. Uniharvest were conducted in the summer of 1978-1979 at Lincoln College, Canterbury. From the first trial, designed to measure dry matter accumulation and changing nutritive value of lupins with time, a maximum dry matter yield of just under 990 kg ha⁻¹ was obtained 150 days after sowing. Peak digestibility was 65.0 per cent and was recorded at 125 days after sowing. Protein concentration was highest in the young plant (28.75 per cent at 45 days after sowing) and declined after this, but highest total nitrogen yield coincided with peak dry matter accumulation. In the second trial, lupins at 60 and 100 plants m⁻² were grazed with weaned lambs at each of four successive stages of the plants' growth - pre-flower, primary flower, secondary flower and green pod (post-flower). Highest dry matter accumulation,at just under 1 000 g m⁻², occurred at the green pod stage, and peak protein concentration of 23.7 per cent at the pre-flower stage (lower density). Digestibility was highest at the pre-flower stage (higher density) and the metabolizable energy concentration was also highest in the pre-flower high density plots. Regrowth of lupins following grazing occurred in plots which had been grazed at the pre- and primary flower stage. Maximum total dry matter accumulation from the first grazing combined with the regrowth was 1 350 g m⁻². Digestibility, protein and metabolizable energy concentration of the regrowth was at acceptable levels for animal growth. From the results obtained, and those of other workers, both in New Zealand and overseas, it is suggested that lupins may have considerable potential as a high quality summer forage crop for grazing of young lambs.

The insect fauna of two Lupinus angustifolius cultivars (Fest and Uniharvest) was surveyed between October, 1978 and March, 1979. The main sampling methods were sweepnetting, use of the D-Vac, and pod removal. Glasshouse experiments were carried out with species showing pest potential (based on the survey). These species were Acyrthosiphon pisum (Harris), Nysius huttoni White, and Sidnia kinbergi (Stalenberg). While A. pisum and S. kinbergi had their pest potential confirmed, N. huttoni did not. Although the aphids Acyrthosiphon kondoi Shinji, Macrosiphum euphorbiae (Thompson) and Myzus persicae (Sulzer) were not studied in the laboratory, field data indicated their pest potential. Insects to show lesser pest potential were Calocoris norvegicus (Gmelin), Hylemyia deceptiva Malloch, and Hylemyia platura (Meigen). Important potential predators or parasites were Austromicromus tasmaniae (Walker), Coccinella undecimpunctata L., Tropiconabis capsiformis (Germar), some Araneae species and single unidentified species of Braconidae and Eulophidae. The cultivar Uniharvest was the preferred host presumably because it does not possess the toxic alkaloids of Fest. However, crop loss through insect damage appeared unimportant compared to the 10% loss of Uniharvest caused by the aphid-borne bean yellow mosaic virus.