This research examined the mechanisms by which temperature, water availability and nitrogen (N) affect the dry matter (DM) yield potential of cocksfoot (Dactylis glomerata L.) dominant pastures. The experiment was a split plot design with main plots of fully irrigated (I) or dryland (D), sub-plots of N fertiliser at 800 kg N/ha in 2003/04; and 1600 kg N/ha in 2004/05 (+N) or 0 kg N/ha (-N). The potential environmental yield of an established 8 year old cocksfoot dominant pasture was 21.9 t DM/ha/y from I+N pastures compared with 9.8 t DM/ha by I-N pastures and 15.1 t DM/ha/y by D+N pastures. The lowest yields were from dryland pastures with no N which produced 7.5 t DM/ha/y in 2003/03 and 5.0 t DM/ha/y in 2004/05. The effect of seasonal temperatures on the DM production, when periods of water stress were excluded, was quantified using thermal time accumulated above a base temperature of 3°C as 7.0 kg DM/°Cd/ha for N fertilised pastures and 3.3 kg DM/°Cd/ha for pastures with no N. The 2.5 t DM/ha difference in yields of D-N pastures in 2003/04 and 2004/05 was the result of the duration, extent and timing of the water stress period. In both years the critical limiting deficit (DL) was calculated as 78 mm from the soil moisture deficit in the 0-0.8 m soil layers. Beyond DL yield decreased at a rate of 1.45%/mm in +N and –N pastures, relative to fully irrigated control pastures. Yields of D+N and D-N pastures were similar during periods of water stress with 0.4±0.1 t/DM/ha produced during the rotation ending 30/12/2003. This was less than from either the I-N (1.2 t DM/ha) or I+N (3.5 t DM/ha) pastures due to the reduction in the amount of photosynthetically active radiation intercepted by the canopies of the dryland pastures. However, in the rotation ending 2/5/2004, after autumn rain alleviated drought conditions, yield of the D+N pasture was 2.1 t DM/ha compared with 1.7 t DM/ha by I+N pastures. The effect of N on yield was described using a nutrition index which showed that as DM yield increased N% in the herbage declined. This is a function of the ratio between metabolic and structural N requirements rather than caused by ontogeny alone. Specific leaf N was determined at two harvests and appeared constant at a given point in time (1.0-1.6 g N/m² leaf). In contrast, specific pseudostem N increased from 0.8-1.0 g N/m² pseudostem at an NNI of 0.4 in –N pastures to 2.6-3.0 g N/m² pseudostem at an NNI of 1.2 in the +N pastures. Differences between the yields of +N and –N pastures were caused by differences in radiation use efficiency (RUE) as determined by the linear relationship (R²=0.76) between RUE and the nitrogen nutrition index (NNI). In this thesis, empirical relationships for the effects of temperature, water availability and N were derived and the physiological mechanisms which underlie these descriptions were identified. These relationships provide clear and simple explanations of the effects of environmental variables on the productivity of cocksfoot based pastures which will enhance understanding of the benefits and limitations of cocksfoot, particularly in dryland farming systems.

This research examined the mechanisms by which temperature, water availability and nitrogen (N) affect the dry matter (DM) yield potential of cocksfoot (Dactylis glomerata L.) dominant pastures. The experiment was a split plot design with main plots of fully irrigated (I) or dryland (D), sub-plots of N fertiliser at 800 kg N/ha in 2003/04; and 1600 kg N/ha in 2004/05 (+N) or 0 kg N/ha (-N). The potential environmental yield of an established 8 year old cocksfoot dominant pasture was 21.9 t DM/ha/y from I+N pastures compared with 9.8 t DM/ha by I-N pastures and 15.1 t DM/ha/y by D+N pastures. The lowest yields were from dryland pastures with no N which produced 7.5 t DM/ha/y in 2003/03 and 5.0 t DM/ha/y in 2004/05. The effect of seasonal temperatures on the DM production, when periods of water stress were excluded, was quantified using thermal time accumulated above a base temperature of 3°C as 7.0 kg DM/°Cd/ha for N fertilised pastures and 3.3 kg DM/°Cd/ha for pastures with no N. The 2.5 t DM/ha difference in yields of D-N pastures in 2003/04 and 2004/05 was the result of the duration, extent and timing of the water stress period. In both years the critical limiting deficit (DL) was calculated as 78 mm from the soil moisture deficit in the 0-0.8 m soil layers. Beyond DL yield decreased at a rate of 1.45%/mm in +N and –N pastures, relative to fully irrigated control pastures. Yields of D+N and D-N pastures were similar during periods of water stress with 0.4±0.1 t/DM/ha produced during the rotation ending 30/12/2003. This was less than from either the I-N (1.2 t DM/ha) or I+N (3.5 t DM/ha) pastures due to the reduction in the amount of photosynthetically active radiation intercepted by the canopies of the dryland pastures. However, in the rotation ending 2/5/2004, after autumn rain alleviated drought conditions, yield of the D+N pasture was 2.1 t DM/ha compared with 1.7 t DM/ha by I+N pastures. The effect of N on yield was described using a nutrition index which showed that as DM yield increased N% in the herbage declined. This is a function of the ratio between metabolic and structural N requirements rather than caused by ontogeny alone. Specific leaf N was determined at two harvests and appeared constant at a given point in time (1.0-1.6 g N/m² leaf). In contrast, specific pseudostem N increased from 0.8-1.0 g N/m² pseudostem at an NNI of 0.4 in –N pastures to 2.6-3.0 g N/m² pseudostem at an NNI of 1.2 in the +N pastures. Differences between the yields of +N and –N pastures were caused by differences in radiation use efficiency (RUE) as determined by the linear relationship (R²=0.76) between RUE and the nitrogen nutrition index (NNI). In this thesis, empirical relationships for the effects of temperature, water availability and N were derived and the physiological mechanisms which underlie these descriptions were identified. These relationships provide clear and simple explanations of the effects of environmental variables on the productivity of cocksfoot based pastures which will enhance understanding of the benefits and limitations of cocksfoot, particularly in dryland farming systems.

Genotypic differences in growth and yield of two common bean (Phaseolus vulgaris L) cultivars to Rhizobium inoculation and management were investigated. In 2003-04, the two bean cultivars (Scylla and T-49) were combined with three inoculant treatments (strains CC 511 and RCR 3644, and a control of no inoculation), two fertiliser levels (0 and 150 kg N ha⁻¹) and two irrigation treatments (irrigated and rainfed). There was no nodulation on either cultivar. To further investigate the symbiotic relationship, 16 rhizobial isolates, including the two used in the first field experiment, were combined with the cultivar Scylla and evaluated in a greenhouse. Subsequently, five Rhizobium isolates were chosen for further field evaluation, based on signs of early nodulation in the greenhouse trial. The second field experiment in 2004-05 combined the five inoculant strains (RCR 3644, UK 2, H 20, PRF 81, PhP 17 and a control) with two bean cultivars (Scylla and T-49). In the greenhouse, nodule number varied from 7 (UK 2) to 347 (H 441) nodules plant⁻¹ at 51 DAS and from 13 (UK 1) to 335 (CIAT 899) nodules plant⁻¹ at 85 DAS. In 2004-05, in the field, nodulation was also variable, ranging between 1 and approximately 70 nodules plant⁻¹, with higher nodules numbers plant⁻¹ being found on cultivar T-49. Of the isolates used in the field, strains H 20, PRF 81 and PhP 17 produced 70, 25 and 12 nodules plant⁻¹ at 70, 40 and 54 DAS respectively. Nodules formed were of various sizes and more than 80 % were pink to dark red in colour denoting the presence of leghaemoglobin and active N fixation. The remaining nodules were either green or white. The importance of selecting an appropriate cultivar for the growing conditions was highlighted in these experiments. Leaf area index, leaf area duration intercepted radiation and final utilisation efficiency were significantly affected by cultivar. In both seasons cv. T-49 reached maturity (dry seed) before Scylla, while unirrigated plants reached green pod maturity seven days before irrigated plants. Plants of cv. Scylla gave a final TDM of 730 g m⁻²; compared to the 530 g m⁻² produced by T-49. The average growth rate was 7.0 and 5.2 g m⁻² day⁻¹ for Scylla and T-49 respectively (2003-04). Plants receiving 150 kg N ha⁻¹ produced 665 g m⁻² TDM which was 12 % more than was produced by unfertilised plants. The application of 150 kg N ha⁻¹ gave an average growth rate of 6.4 g m⁻² day⁻¹ compared to 5.7 g m⁻² day⁻¹ from plants with no N. Inoculation in the field had no significant effect on TDM in both seasons. Temperature affected growth and DM accumulation. Accumulated DM was highly dependent on cumulative intercepted PAR. Air temperatures below the base temperature (10 °C) affected growth in 2004-05, resulting in plants accumulating just 0.24 g DM MJ⁻¹ PAR during early growth. This increased to 2.26 g DM MJ⁻¹ PAR when the temperature was increased above the base temperature. There was a strong relationship between LAI and intercepted PAR. A LAI of 4.0-4.5 was required to intercept 90-95 % of incident solar radiation. Cultivar significantly (p < 0.001) affected radiation use efficiency (RUE). Scylla had a RUE of 1.02 g DM MJ⁻¹ PAR compared to T-49 at 1.18 g DM MJ⁻¹ PAR. Seed yield was significantly (p < 0.001) affected by cultivar and fertiliser application. Cultivar Scylla produced 467 g m⁻² which was 76 % more than T-49, while a 12 % increase in seed yield was observed in N fertilised plants over unfertilised plants. Only cultivar significantly affected HI, while the yield components that had the greatest effect on seed yield were hundred seed weight and pods plant⁻¹. Inoculation significantly (p< 0.05) affected 100 seed weight (2004-05). Plants inoculated with strain H 20 had the highest 100 seed weight at 25.2 g with cv. Scylla producing larger seeds than T-49. The belief that local environmental conditions play a major role on field survival of bacteria, led to the use of PCR methods to identify field nodulating organisms. Amplification of genomic DNA from parent isolates using primers fC and rD generated a single band for each isolate. Isolates were identified to the species level as either Rhizobium or Agrobacterium, using the highly conserved internally transcribed spacer (ITS) region and are known to nodulate common bean. The DNA extracted from the isolates recovered from nodules of field grown beans gave multiple bands with primers fC and rD. Five distinct banding patterns were observed. All of these were different from those of parent isolates. Sequencing of the 16S rRNA demonstrated that nodules of field grown beans in Canterbury were inhabited by Pseudomonads either alone or in association with other root nodulating organisms. The inability to identify the inoculant strains in nodules of field grown beans does not rule out their infection and nodulating function in the cultivars used. The results suggest the possibility of both Rhizobium and Pseudomonads cohabiting in the nodules of field grown beans. The aggressive nature of Pseudomonads on artificial media, possibly out competing the inoculant rhizobia is proposed, leading to the inability to identify the inoculant strain from the nodules of the field grown beans by PCR methods. The need to identify the nodule forming or nodule inhabiting bacteria in the nodules is necessary to classify the importance of these organisms and their economic benefit to agricultural production. This study also underlines the importance of using PCR methods to gain valuable insights into the ecological behaviour of Rhizobium inoculants and nodule inhabiting organisms.

CoordenaÃÃo de AperfeÃoamento de Pessoal de NÃvel Superior / Avaliou-se as trocas gasosas, o fluxo de biomassa e as caracterÃsticas estruturais do capim-aruana pastejado sob lotaÃÃo rotativa por ovinos Morada Nova submetidos a quatro nÃveis de suplementaÃÃo concentrada (0,0; 0,6; 1,2 e 1,8% do PC), com perÃodo de ocupaÃÃo de quatro dias e descanso quando da interceptaÃÃo de 90% da radiaÃÃo fotossinteticamente ativa (IRFA), num delineamento inteiramente casualizado, num arranjo em parcelas subdivididas, sendo os nÃveis de suplementaÃÃo as parcelas e os ciclos de pastejo as subparcelas. A condiÃÃo de pÃs-pastejo consistiu em um Ãndice de Ãrea foliar de 1,0. Constatou-se efeito linear crescente para a taxa de alongamento das hastes e taxa de aparecimento foliar em funÃÃo dos nÃveis de suplementaÃÃo. As taxas de produÃÃo e acÃmulo de forragem foram incrementadas pelos nÃveis de suplementaÃÃo. As variÃveis das trocas gasosas das folhas recÃm-expandidas: transpiraÃÃo, condutÃncia estomÃtica, taxa de fotossÃntese foliar, mostraram efeito somente de ciclo de pastejo. NÃo houve nenhum efeito significativo sobre as variÃveis das trocas gasosas das folhas emergentes. Os componentes da biomassa pÃs-pastejo registraram efeito significativo da interaÃÃo nÃveis de suplementaÃÃo e ciclos de pastejo para biomassa de forragem verde residual (BFVr), lÃmina foliar verde residual (BLVr), colmo verde residual (BCVr) e relaÃÃo folha:colmo (F_Cr) residual, estas variÃveis, com exceÃÃo da F_Cr apresentaram efeito quadrÃtico nos ciclos dois e trÃs, confirmando o efeito substitutivo, a partir do nÃvel de suplementaÃÃo de 0,86%, com o mÃnimo estimado de 257,53 Kgâha-Ââano-1, para a BLVr. A relaÃÃo material vivo_ material morto residual apresentou efeito linear crescente em funÃÃo dos nÃveis de suplementaÃÃo. A altura do pasto e a densidade populacional de perfilhos residual foram afetadas somente pelos ciclos de pastejo. A biomassa de forragem verde (BFV) e de lÃmina foliar verde (BLV) apresentaram efeito da interaÃÃo nÃveis de suplementaÃÃo e ciclos de pastejo. A BLV registrou efeito quadrÃtico para os ciclos dois e trÃs, onde os maiores valores foram de 1489,47, no nÃvel de 0,62, do segundo ciclo e de 1554,95 Kgâha-Ââano-1no ciclo trÃs com nÃvel de suplementaÃÃo de 0,14 %. As trocas gasosas nÃo sÃo afetadas pelos nÃveis de suplementaÃÃo. As caracterÃsticas morfogÃnicas do pasto de capim-aruana sÃo afetadas linearmente pelos nÃveis crescentes de suplementaÃÃo. Os componentes estruturais do pasto de capim-aruana sÃo afetados pelos nÃveis crescentes de suplementaÃÃo concentrada, registrando efeito substitutivo do suplemento. / This study evaluated the gas exchange, the biomass flow and the structural characteristics of Panicum maximum cv. Aruana grazed by Morada Nova sheep under rotational stocking and supplied with four concentrate supplementation levels (0.0; 0.6; 1.2 and 1.8% body weight), using four day-grazing periods and rest when the interception of 90% of the photosynthetic active radiation (PAR). The experiment was a completely randomized split-plot design with supplementation levels as plots and grazing cycles as subplots. The post-grazing condition consisted of a leaf area index of 1.0. We observed an increasing linear effect for stem elongation rate and leaf appearance rate according to the supplementation levels. The production and accumulation rates of forage also increased according to the supplementation levels. The different grazing cycles affected the variables of gas exchange of the recently expanded leaves, namely transpiration, stomatal conductance, leaf photosynthesis rate. In turn, no significant effect was detected for the gas exchange variables of emergent leaves. Biomass components during the post-grazing period showed a significant effect of the interaction between supplementation levels and grazing cycles for residual green forage biomass (BFVr), residual green leaf biomass (BLVr), residual green stem biomass (BCVr) and residual leaf:stem ratio (F_Cr). All these variables, with the exception of F_Cr, presented a quadratic effect in the cycles two and three, confirming the substitutive effect from the supplementation level of 0.86%, with the minimum estimated at 257.53 Kgâha-Ââyear-1, for BLVr. The ratio between living and senescent material showed an increasing linear response as a function of the supplementation levels. The pasture height and the residual tiller population density were affected only by grazing cycles. Green forage biomass (BFV) and green leaf biomass (BLV) presented a quadratic effect of the supplementation levels and grazing cycles. BLV showed a quadratic effect for the cycles two and three, where the higher values were 1489.47, at the level of 0.62, in the second cycle, and 1554.95 Kgâha-Ââyear-1 in the third cycle with a supplementation level of 0.14%. Gas exchange was not influenced by the supplementation levels. Morphogenetic traits of the aruana grass pasture are linearly affected by increasing levels of supplementation. The structural components of the aruana grass pasture are influenced by the increasing levels of concentrate supplementation, with a substitutive effect of the supplement.