Search

This research examined the potential of Lupinus polyphyllus L. (also known in New Zealand as perennial lupin and Russell lupin) as a productive pasture component for high-country pastures. To do this, a combination of on-farm studies, controlled field experiments and glasshouse experiments were carried out. Throughout the thesis, lucerne (Medicago sativa L.) was used as a control. The use of perennial lupin as a suitable feed for a commercial Merino ewe flock was evaluated in an on-farm study at Sawdon Station, Lake Tekapo. The performance of ewes and lambs grazing a lupin-based pasture was compared with a control flock that was predominantly grazing lucerne. The lupin pasture grew rapidly during the spring period and average pasture cover increased by about 4500 kg of dry matter (DM)/ha under a stocking rate of 15-17 stock units/ha. At tailing in December 2012 and 2013, lambing averaged 111% and ewes averaged 58 kg for the lupin pasture, and 105% and 62 kg for the control flock, while lambs averaged 19 kg for both mobs. At weaning in February 2013, ewes and lambs on the lupin pasture averaged 58 kg and 28 kg compared with 63 kg and 31 kg for the control flock, respectively. During autumn, ewes on the lupin pasture gained 3.8 kg compared with 5.5 kg for the control flock before mating in May. In September, the ewes were shorn and wool averaged 4.62 kg/ewe for the lupin mob and 4.92 kg/ewe for the control flock, with a mean fibre diameter of 18.5 µm. A selection of L. polyphyllus samples were subjected to wet chemistry analyses, which was later used to calibrate a near infrared spectroscopy analyser and predict the nutritional composition of lupin material. Lupin lamina was consistently high in metabolisable energy (11.3-12.6 MJME/kg DM) and crude protein (24-33%) throughout the growing season. Under controlled experimental conditions at Lincoln University, sheep liveweight gain from a dryland mixed cocksfoot (Dactylis glomerata L.)-lupin pasture was 810 kg/ha compared with 1227 kg/ha from dryland lucerne over a 15-month period. This difference was primarily driven by the amount of palatable feed consumed by the sheep, which was 7330 kg DM/ha for cocksfoot-lupin compared with 10,922 kg DM/ha for lucerne. The abundance of lupin in the mixed pasture declined from 40% of the total herbage offered, during the first autumn after sowing (March to May, 2014), to 22% of total herbage offered in the first full growth season (August 2014 to May 2015). This meant that at least 60% of the diet of these sheep was comprised of cocksfoot. In late spring/summer, the growth rate of cocksfoot-lupin pastures slowed to 20 kg DM/ha/day compared with 43 kg DM/ha/day for lucerne. Despite these differences in feed consumption and pasture composition, the feed conversion efficiency of sheep liveweight gain was the same for both pasture types and averaged 131 g of liveweight gain/kg of DM consumed from August 2014 to May 2015. The greater sheep liveweight gain and total feed consumed on lucerne compared with cocksfoot-lupin was associated with a greater DM yield through improved water-use efficiency for lucerne. During the 2014/2015 growth season, the cocksfoot-lupin pasture produced 6612 kg DM/ha compared with 8122 kg DM/ha for lucerne before a soil moisture deficit started to limit pasture growth in summer 2014/15. To produce this yield, lucerne used 404 mm of water whereas cocksfoot-lupin used 360 mm of water. This equated to a water use efficiency of 20.9 kg DM/ha/mm for lucerne and 18.6 kg DM/ha/mm for cocksfoot-lupin. The ‘Monteith framework’ was successfully applied to both pasture types and provided an accurate description of water extraction patterns to a depth of 165 cm. Lucerne extracted an additional 53 mm of moisture than cocksfoot-lupin from the soil profile between 55 and 155 cm. These results helped to explain the superior sheep liveweight gain on lucerne. Lupinus polyphyllus plants were heavily nodulated at 10 field sites sampled across the South Island. Twenty-two bacterial isolates formed functional nodules on L. polyphyllus, which indicated that rhizobia, capacble of nodulating L. polyphyllus, were present across a wide range of sites in the South Island. Gene sequences identified all 22 isolates and the Group G commercial inoculant as Bradyrhizobium sp. Eleven isolates and the Group G inoculant were tested for their effectiveness on growth of L. polyphyllus under glasshouse and field conditions. Plants were grown in a high-country soil under glasshouse conditions and all plants nodulated regardless of inoculum treatment. However, plant growth was variable and further quantification was required. Field studies showed that inoculation was beneficial for the growth of L. polyphyllus seedlings under high-country conditions. Therefore, despite the presence of effective rhizobia throughout the South Island, inoculation is recommended for newly established stands. Both L. polyphyllus and lucerne responded to increasing levels of soil nitrate by reducing N fixation. Application of potassium nitrate at rates up to 600 kg N/ha had little effect on the growth or herbage N concentration of L. polyphyllus or lucerne. However, the proportion of N derived from soil (%Ndfs) increased consistently with fertiliser applications between 50 and 600 kg N/ha. For both species, %Ndfs was highest at a fertiliser rate of 600 kg N/ha, where %Ndfs was 62% for L. polyphyllus and 74% for lucerne. This thesis confirmed that L. polyphyllus is a suitable forage option for dryland, high-country, farmers. Rather than replacing lucerne, L. polyphyllus should be viewed as a complementary species that can considerably improve the productivity of areas of lower soil fertility. In this context, L. polyphyllus is best suited to areas that are often uneconomical to develop for traditional legumes. Farmers will need to carefully manage L. polyphyllus during establishment, with a focus on seedbed preparation and careful grazing management during the first season. However, once established, L. polyphyllus will likely remain as a persistent dryland species that will produce significant amounts of spring forage on high-country farms.

Soil acidity and associated soil aluminium (Al) toxicity severely restrict the establishment, yield and persistence of legumes in New Zealand high and hill country pastures. There is an urgent need to identify soils which are most susceptible to Al toxicity and to determine what factors are involved in regulating soil extractable Al concentrations. This study investigated the relationship between soil chemical, physical and environmental variables and soil extractable Al for New Zealand soils in national, catchment and rhizosphere contexts. The relationship between soil chemical, physical and environment variables and soil extractable AlKCl were investigated using the National Soils Database (NSD). Base saturation (BS), soil pHH2O, cation exchange capacity (CEC), total nitrogen, total carbon and soil order were strongly associated with AlKCl concentrations and relationships differed among the depth zones (0-20 cm, 20-50 cm and 50-120 cm). Soil acidity and high CEC contributed to high AlKCl concentrations, whereas a high BS and total C had a negative effect. Total N decreased with increasing AlKCl in the top 20 cm, likely as a response to Al toxicity induced limitations on biological N fixation by pasture legumes. An AlKCl concentration > 1.0 cmolc/kg, which can be toxic to sensitive plants, occurred across a pHH2O range of 3.8-6.4. Brown Soils and Podzols had the highest mean AlKCl concentrations across all depths and are likely to be more susceptible to Al toxicity. A soil survey in the Ashburton Lakes catchment was conducted to determine which soils are most susceptible to Al toxicity and to identify which factors drive soil pHH2O and extractable AlCaCl2 at a landscape scale. Depth in the soil profile was the strongest explanatory variable for pHH2O and AlCaCl2. Soil pHH2O increased with depth and AlCaCl2 declined. Rainfall and age were significant factors for AlCaCl2, however, there were no systematic patterns of an increase in AlCaCl2 with increasing rainfall and soil age. Differences in pHH2O among landform types were found, in contrast to no difference in AlCaCl2 concentrations. The soil pHH2O ranged from 4.7-6.0 and AlCaCl2 concentrations from 1.2 mg kg-1 to 39.1 mg kg-1, with a mean of 7.8 mg kg-1. Maps of soil pHH2O and AlCaCl2 in the 20 cm depth zone were constructed using the rules established by decision trees. Distinct areas in the landscape were identified which had higher concentrations of AlCaCl2. Higher AlCaCl2 concentrations were found at the wettest sites in the catchment (≥1266 mm), areas that seem to mirror those that were identified as most acidic in the soil pHH2O map. The growth response and nutrient uptake of legumes (Medicago sativa L. and Trifolium ambiguum L.) as bioindicators of Al toxicity were assessed in a range of acidic soils in a glasshouse experiment. Soil extractable AlCaCl2 concentration was strongly associated with lucerne shoot yields. Lucerne shoot yield increased one to six fold with lime application, particularly between the L0P0 and 2 t lime ha-1 treatments. Yield increases were strongly associated with declines in the soil extractable AlCaCl2 concentration to below toxic levels (≤2.5 mg kg-1) and clearly demonstrated the severe plant growth restriction of Al toxicity in these high country soils. On most soils lucerne shoot yields responded more to lime than P applied. In contrast, Caucasian clover (CC) shoot yields were not affected by soil extractable Al concentration, with more consistent yields across the range of pHH2O (5.0-7.5) and the P rates applied than lucerne. This study clearly highlights the potential importance of CC use in the high country, where the growth of more sensitive species such as lucerne is restricted by Al toxicity. The plant effect on Al mobilisation and immobilisation at the root-soil interface and the effects of pHH2O were investigated in a rhizobox experiment with legumes (Lupinus polyphyllus L. and Medicago sativa L.) in an acidic high country soil. The pHH2O was more acidic (0.1- 0.3 pH units lower) and AlCacl2 concentrations were higher (0.5 mg kg-1 and 5.4 mg kg-1) in the rhizosphere of lupin plants compared to the bulk soil. Lucerne plants had a similar soil pHH2O between the bulk and rhizosphere. Hot water extractable organic carbon levels appeared to be consistently higher in the rhizosphere and seemed to increase at the highest lime application. DGT data showed increased mobilisation of Al at the root tip of lupin and depletion along the root axis, indicative of previous removal of Al by the plant from the soil. This is the first study that has shown, in high resolution using DGT and LA-ICP-MS analysis, distinct patterns of soil Al mobilisation induced by the roots of important pasture species. A laboratory investigation was conducted to determine if changing the molarity and extraction time of the standard CaCl2 and KCl soil Al tests altered the Al concentrations extracted. Overall, the Al concentration extracted by the KCl standard test was 16 times higher than the CaCl2 across all soils. The effect of molarity and extraction time on the Al concentrations extracted differed among the five soils tested for the two extraction methods. For the CaCl2 test, the extractable Al concentration increased (P0.05) with a further increase in molarity. For two of the soils, the Al extracted was significantly affected by the extraction time, however, the results were contrasting. On most soils the interaction of molarity and extraction time extracted significantly different concentrations of AlKCl. These findings suggest that the Al concentrations measured by the two extraction methods are affected by specific soil properties in the topsoil related to soil order. This research has identified key variables driving soil extractable Al concentrations in a suite of New Zealand soils, and at different scales. Legume species responded differently to soil extractable AlCaCl2 and influenced pH and the amount of soil Al in the rhizosphere. Soil properties of different soil orders were also found to affect the amount of Al extracted from the soil by the two CaCl2 and KCl extraction methods. The knowledge generated from this thesis has identified specific sets of conditions (environmental, soil chemical and soil order) that have higher concentrations of extractable Al and therefore areas most likely to be susceptible to Al toxicity in New Zealand.

Coumestrol in lucerne can reduce the ovulation rate of ewes, which lowers lambing rates. This thesis isolated the factors with potential to increase the coumestrol levels of lucerne. The management of ewes during the mating season to avoid suppressed ovulation was also investigated. This enabled management strategies to be developed to mitigate the risk of depressed ovulation rate in ewes grazing lucerne during the mating season. Agronomic factors investigated were lucerne cultivar, cutting frequency, development stage, aphid herbivory, fungal infection, water stress, and fungicide application. Fungal pathogens explained most of the differences in coumestrol with a moderate relationship between fungal damage and coumestrol content in field and glasshouse experiments. Lucerne inoculated with Stemphylium sp. had high coumestrol levels, with 169 ± 25.1 mg/kg DM measured in entire shoots, compared with 3.4 ± 0.84 mg/kg in control plants. Pea aphids were a minor cause of a coumestrol response, with an increase from 2.4 ± 0.39 mg/kg DM to 5.3 ± 0.65 mg/kg DM measured in a glasshouse experiment. Flowering did not produce a coumestrol response, with simultaneous increases observed in vegetative and flowering plants. Wilted lucerne had a coumestrol content of 1.3 ± 0.43 mg/kg DM which was not different to well-watered lucerne (2.2 ± 0.59 mg/kg DM) and only increased to 3.0 ± 0.57 mg/kg DM with further stress. Despite fungal pathogens causing the main coumestrol response, carbendazim fungicidal treatments predicted to decrease coumestrol accumulation were ineffective. In leaves inoculated with stemphylium a newer cultivar ‘Stamina 5’ had higher coumestrol (396 ± 82.4 mg/kg DM) than ‘Wairau’ (143 ± 35.6 mg/kg DM), a 40 year old industry standard. In the field, five cultivars currently on the market did not differ in coumestrol content throughout a growing season and all reached at-risk levels (> 25 mg/kg DM) with an average content of 56.2 ± 3.24 mg/kg DM by late-May 2015. Removal of herbage was an effective method to reduce coumestrol as regrowth material typically had lower levels than the prior herbage. Based on the field data, a model which used relative humidity and rainfall terms was created to predict when lucerne was likely to have heightened coumestrol. From this model a risk assessment analysis was created for a range of New Zealand locations. The assessment showed that four week old lucerne regrowth was at lower risk of elevated coumestrol than the standard six week regrowth which ewes are typically grazed on. Blenheim was least likely of the four locations to have at-risk levels of coumestrol and in all locations except Napier, the risk of high coumestrol increased as the autumn progressed. This would have greatest effect in Otago/Southland, where mating is concentrated late in the season. Indirect prediction of coumestrol content was also identified through ewe lambs that showed increased growth of teats and udder protrusion as an indicator of the crop oestrogenicity. Ideally, high quality forage provided by lucerne is beneficial for live weight and thus reproductive performance, but grazing must avoid coumestrol-induced impairment of ovulation rates. A grazing experiment removed ewes at different intervals prior to a CIDR-induced ovulation. This experiment demonstrated that the risk of impaired lambing performance due to moderately oestrogenic lucerne consumption decreased with time on grass prior to ovulation. Removal of ewes two weeks prior to ovulation sufficed to mitigate the risk of decreased lambing performance in this situation. The main results of this thesis indicate coumestrol levels are elevated by fungal infection of lucerne, but grazing management can be utilised to mitigate the effects on ewe reproductive performance.

The productivity of a lucerne monoculture, lucerne/brome and lucerne/cocksfoot binary mixtures grown on shallow soils at Ashley Dene, Canterbury, were compared under grazing over four years. Pastures were grazed with separate mobs of ewes & lambs during the spring, weaned lambs during the summer, and hoggets during autumn. The live weight gain of sheep was the ultimate measurement of pasture performance. In turn, this was related to seasonal pasture yield and pasture composition. The live weight production of the lucerne monoculture averaged 28% more than the binary mixtures, which were not different from each other. During the first year annual live weight gain was 780 (±26.2) kg/ha for all treatments, because lucerne dominated their composition. Thereafter, lucerne made up >70% of pasture yield in the lucerne treatment and ~37% in the binary mixtures. Live weight gain was 550-800 kg/ha on the lucerne treatment and 400-550 kg/ha on the binary mixtures over the next three years. Annual pasture yields were similar among treatments each year. These ranged from 6.3-10.8 (±0.28) t DM/ha with soil water availability during the growing season being a strong predictor of yield. The growth response to thermal time, while it appeared soil water was not limiting, averaged 6.1 (±0.21) kg DM/ha/°C day during the first two years and 4.3 kg DM/ha/°C day during the last two. The difference among years was related to higher pre-grazing pasture covers during the first two years which probably increased interception of photosynthetically active radiation. The onset of a soil water deficit, which limited pasture growth, was in mid-December during the first two years and in early/mid-November during the final two. This occurred at comparable times for each pasture treatment suggesting soil water accessibility was independent of this, on these shallow soils. This is consistent with 184 mm of soil water accessibility. The higher live weight gain of the lucerne treatment eventuated through up to 30% higher metabolisable energy intake compared with the binary mixtures. The shoot structure of lucerne appeared to facilitate enhanced discrimination between more and less palatable portions during grazing relative to sown grass species. This may be complemented by the lower levels of neutral detergent fibre in lucerne which results in faster digestion and therefore increased pasture intake.

This dissertation compared lucerne dry matter production from 10 genotypes in irrigated conditions at Lincoln University. First year yields averaged 10 t DM/ha for the establishment year with no differences found among culitvars.

This research focused on perennial legumes for the low fertility regions of the New Zealand high country. The aim was to improve production of legume species in these regions. A six-year field experiment at the Lees Valley evaluated the production of lucerne (Medicago sativa L.), Caucasian clover (Trifolium ambiguum M. Bieb.), white clover (T. repens L.), and red clover (T. pratense L.). Low soil pH and phosphorus deficiency were addressed by lime and P fertiliser application, but lucerne persistence was low at this site. Caucasian clover was the most persistent legume with the highest sown species component (93%) and yield (7.8 t DM/ha) in last season. Lucerne yielded 4.2 t DM/ha in Year 1, but this decreased to 2.7 t DM/ha in Year 3, and 1.0 t DM/ha in Year 6. Causes of these extreme responses were examined in a series of pot experiments. In the Lees Valley soil, 85% of lucerne seedlings died, when no lime was applied (pH 5.4, Al> 6.0 mg/kg) compared with 40% of Caucasian clover seedlings. The 2 t/ha lime application, increased lucerne seedling survival to 80% (pH 6.1, Al< 0.03 mg/kg). A similar survival rate was achieved for Caucasian clover with 1.0 t lime/ha. Results showed low pH and high Al content of the Lees Valley soil suppressed the growth of fine roots and nodulation of lucerne, more than Caucasian clover. However, the level of difference in Al tolerance between the two species could not fully explain the field experiment results at the Lees Valley. The main limitation for lucerne persistence was nitrogen deficiency, as a consequence of nodulation failure due to the Al toxicity. Three field experiments were conducted at Glenmore Station and Ashley Dene, to further confirm results. Russell lupin (Lupinus polyphyllus L.) was included in these later experiments as a potential Al tolerant legume. The percentage of nodulated plants for Russell lupin was constantly over 75%, over two years of the field experiments at Glenmore station and Ashley Dene. In contrast, for lucerne and Caucasian clover, nodulation decreased to zero and 25%, respectively over time in the high Al soil > 5 mg/kg at Glenmore station, but not under moderate levels of Al (ca. 3-4 mg/kg soil) at Ashley Dene. Nodule persistence of lupin was a key factor for plant survival in the high Al soil. At Glenmore station, Russell lupin was successfully established into resident vegetation and yielded over 10 t DM/ha in Year 2. Its vigorous growth was supported by consistent nodulation (75%), in the high aluminium content soil. This highlighted its potential to provide a nitrogen fixing perennial legume for environment that is unsuitable for lucerne. Caucasian clover establishment was low in both field experiments, but yielded 4 t DM/ha in Year 2 at Glenmore station. At Ashley Dene the nodulated lucerne plants produced 8.0 t DM/ha, compared with 2.0 t DM/ha in un-nodulated plants. The ¹⁵N natural abundance method showed ca. 70% of the nitrogen in nodulated plants was derived from BNF. This meant 122 kg N/ha of inoculated lucerne was fixed within six months. Thus lucerne was highly reliant on nitrogen fixation in an N deficient soil. Surface applied lime rates >4 t/ha, increased the top-soil pH from 5.2 to 5.8 and decreased exchangeable Al levels to less than 3 mg/kg in 0-75 mm of soil depth, in the Lees Valley and at Glenmore Station. However, toxic Al levels in deeper soil horizons and its variability compromised the efficiency of surface lime application to enable lucerne persistence at both sites. The measurable effects of lime were less obvious in 75-150 mm of soil depth. Similar changes in soil pH and Al levels were measured at Ashley Dene but only 2 t lime/ha was required. Combined results from the field and pot experiments suggested the main limiting for lucerne and Caucasian clover was nitrogen deficiency that should be provided from BNF in these low N and high Al soils. Therefore the fate of applied rhizobia inoculants, and the possibility of competition with any naturalized rhizobia strains inhabiting lucerne and Caucasian clover nodules was assessed. Eight naturalized strains of Sinorhizobium meliloti were identified from lucerne nodules grown in different high country regions. This was the first identification report of any naturalized S. meliloti occupying lucerne nodules in New Zealand. Soil pH and related Al levels affected the contribution of rhizobia strains to occupy lucerne nodules. At soil pH of 5.5 (2.8 mg Al/kg), 50% of the recovered isolates were the commercial strain. This proportion decreased to 6%, as the commercial genotype had been replaced by naturalized S. meliloti strains when the soil pH increased to 6.7 (Al< 0.3 mg/kg). Further studies are therefore required to evaluate competitive ability, nodulation and nitrogen fixation rate of these identified naturalized strains, compared with the commercial strain. In contrast for Caucasian clover, the applied commercial strain was the only identified Rhizobium leguminosarum from nodules, irrespective of sites. This indicated high specificity of Caucasian clover for rhizobia symbiont to nodulate the roots.

The objective of this thesis was to compare herbage dry matter (DM) production, nutritive value and water use efficiency of simple perennial ryegrass/white clover and tall fescue-white clover pastures to diverse pastures where additional grasses, legumes and herbs were added to simple grass-clover mixture. Three experiments were conducted in Canterbury, New Zealand with all pastures irrigated and grazed by dairy cows. The first study, conducted at a paddock scale, measured annual and seasonal herbage DM production, botanical composition and nutritive value of two species mixtures of perennial ryegrass (standard and high sugar) and tall fescue sown with white clover compared with more diverse mixtures where additional herbs (chicory and plantain), legumes (lucerne or red clover) and grasses (prairie grass) were added to the two species mixtures. Averaged over 2 years, annual herbage DM production was 1.6 t DM/ha greater in diverse (16.8 t DM/ha) than simple (15.2 t DM/ha) pastures, primarily reflecting greater DM production in summer. Diverse pastures had lower metabolisable energy (ME) (12.0 vs 12.2 MJ ME/kg DM) and neutral detergent fibre (301 vs 368 g/kg DM) content than simple pastures, although the total ME produced per year was greater in diverse than simple pastures (202 vs 185 GJ ME/ha). Ryegrass-based pastures had higher annual DM production (16.8 t DM/ha) than tall fescue-based pastures in the first (14.5 t DM/ha) but not second year. The second study, conducted at a small plot scale, measured annual and seasonal herbage DM production, botanical composition and nutritive value of simple and diverse pasture mixtures grazed by dairy cows subjected to full and partial irrigation. Measurements were made over two years for a simple perennial ryegrass/white clover pasture (S) and pastures with additional legumes (red clover and lucerne, SL), herbs (chicory and plantain, SH), grasses (prairie grass and timothy, SG), herbs and legumes (SLH), or herbs, legumes and grasses (SLHG) added to the simple pasture. In the partial irrigation treatment, no irrigation was applied for a 2 month period from 14 January 2011 to 10 March 2011 (Year 1) and from 7 January to 16 March 2012 (Year 2). Averaged over two years, annual herbage DM production was greater where additional legumes were added to mixtures than where additional herbs or grasses were added or in the simple mixture (16.5, 16.1, 15.1, 14.5, 14.1, 13.6 t DM/ha for SLHG, SLH, SL, SH, S and SG, respectively). The decline in DM yield associated with partial compared to full irrigation was lower in SL (10%) and SLH (14%) and SLHG (15%) than SG (19%) and S (26%) over the two year period. It was concluded that this reflected greater growth of the tap rooted legumes lucerne and red clover during the period of irrigation restriction. All pasture mixtures had similar ME content (11.1 to 11.5 MJ ME/kg DM) but mixtures containing additional legumes (SL, SLH and SHLG) had higher crude protein content (210 to 215g/kg DM) than mixtures containing additional grasses or the simple mixture (184 to 195 g/kg DM). The third study measured water use in the small plot study outlined in study two in order to investigate the production differences between full and partially irrigated pastures. Neutron probe tubes were inserted to a depth of 2.5 m and water use measured. Water use was greater in mixtures containing additional legumes (SL, SLH and SLHG, 689 mm to 705 mm) than the mixture containing additional grasses (SG, 680 mm) or the simple mixture (S, 670 mm). However, the effect was relatively small (range 14 mm to 35 mm) and was tightly linked to the summer period. In mixtures containing additional legumes (SL, SLH and SLHG), water was extracted to greater depths (0-2m) than the mixtures containing additional grasses or the simple mixture (S, SG; 0-1m). Mixtures with additional herbs (SH) extracted water to 0-1.5m soil depth. Water use efficiency (WUE) was greater in the mixtures containing additional legumes and additional herbs (SH, SL, SLH and SLHG, 18.6 to 21.1 kg DM/ha/mm) than additional grasses (SG, 17.5 kg DM/ha/mm) or the simple mixtures (S, 18.3 kg DM/ha/mm). In conclusion, the DM production and nutritive value of diverse pastures was similar or greater than that of standard perennial ryegrass/white clover pastures or tall fescue-white clover pastures. Under water restriction, DM production was less affected in mixtures containing the tap rooted legumes red clover and lucerne. Combined with the environmental benefits of diverse pastures (e.g. reduced urinary N excretion) demonstrated in other studies, it is concluded that diverse pastures are a promising alternative to perennial ryegrass-white clover pastures to deliver high production, with lower environmental implications, in dairy systems.

Lucerne (Medicago sativa L.) is an important legume for New Zealand dryland farms. To ensure the productivity and persistence of lucerne stands rotational grazing is recommended. However, over spring ewes and lambs are typically set-stocked for ease of management. Two grazing experiments were therefore imposed on lucerne stands to quantify the effect of spring grazing management treatments on the animal and plant production, particularly the persistency of the lucerne stand through the perennial reserves composition and quantity. The overall liveweight gains were similar between treatments with rotational having 894 kg LW/ha, semi set-stocked 881 kg LW/ha, and set-stocked having 890 kg LW/ha. The rotational treatments total dry matter production was 12,106 kg DM/ha. Plant population was the highest under set-stocked treatments with 206 plants/m², but it also had the biggest decline, to 134 plants/m² in May. The rotational treatments had the highest total root weights across all treatments, with 4,682 kg/ha, compared to semi set-stocked and set-stocked with 3,443 and 3,989 kg/ha, respectively. Although liveweight gains between treatments were similar, the carbon and nitrogen reserves in the rotational treatments were higher than semi set-stocked and set-stocked treatments. This implies that the persistence of the rotational stand will be greater than the semi set-stocked and set-stocked stands. However, results suggest that set-stocking for a short period of time (e.g. a month) is possible over spring provided that lucerne growth exceeds demand.

Lucerne (Medicago sativa L.) is the highest yielding temperate forage legume, and fixes as much as 360 kg N/ha/year. Its establishment in dryland Canterbury depends on understanding critical factors such as sowing time and inoculation techniques that increase chances of seedling survival and stand persistence. Measurements occurred for a complete growing season after sowing ‘Stamina 5’ lucerne inoculated with peat, lime coating, ALOSCA ® granules and a bare seed control on 4 October, 4 November, 2 December 2010 and 10 January and 7 February 2011 at Lincoln University Canterbury. No emergence differences were observed from the different inoculants. Coated seed produced a higher initial plant population of 331 plants/m² which was 68% of the 490 seeds/m² sown. It also had the highest population 12 months later although seedling numbers had declined by 27%. Regardless of differences in plant population per treatment, dry matter (DM) yields were not different. Accumulated crop DM yields were however different among sowing dates and decreased with each successive sowing. SD 1 (4 October) gave the highest yield of 15 t/ha whilst SD 5 (10 January) yielded only 2.5 t/ha. Lower DM yields in the later sowing dates were explained by less intercepted light and reduced radiation use efficiency (RUE) probably due to changes in DM partitioning from shoots to roots. Specifically, SD 5 had a longer phyllochron (65 °Cd per node) than SD 1 (44.6 °Cd/node) at a base temperature of 5 oC. No differences were observed in canopy architecture for the different sowing dates as shown by a common extinction coefficient of 0.75. RUE for shoot production (RUEshoot) in regrowth crops of all sowing dates was 1.1±0.065 g DM MJ of intercepted photosynthetically active radiation (PARᵢ). This was 18% higher than the 0.9±0.033 g DM/MJ in seedling crops. Crops from the latest sowing grew 1.84±0.26 kg/ °Cd compared with 3.84±0.22 kg/ °Cd for the earliest sowing. Thermal time accumulation to 50% buds visible (Tt0-bv) decreased from 686 °Cd to 651 °Cd as the photoperiod lengthened in summer from 13 to 15 hours for SD 1 and SD 2. Genotypic characterisation of bacteria isolated from inoculated root nodules and its identification using PCR and 16S methods generated bands for each isolate. These were sequenced and identified to genus level revealing the presence of seven common genotypes. Amongst these, Rhizobium tibeticum was the most dominant species with an occurrence of 14.5%. Ensifer meliloti, the rhizobia responsible for nodulation and subsequently N₂ fixation in lucerne was the second most dominant with an occurrence of 14%. It was also found present in all inoculant carriers characterised but not the bare seed control. These results suggest dominance and effectiveness of resident soil bacteria over introduced Ensifer meliloti in nodulating lucerne. The study emphasises the importance of sowing early in spring or summer to take advantage of the longer growing season and the use of the summer fallow to supplement soil moisture for successful lucerne establishment.

The annual and seasonal water use efficiency of six pasture combinations were calculated from the ‘MaxClover’ grazing experiment at Lincoln University, Canterbury. Pastures had been established for seven years and were grazed by best mangement practises for each species combination. Measurements from this study are from individual plots of six replicates of cocksfoot (CF)/Subterranean clover (Sub), CF/Balansa clover (Bal), CF/White clover (Wc), CF/Caucasian clover (Cc), ryegrass (RG)/Wc and lucerne (Luc). Dry matter measurements of yield, botanical composition and herbage quality were assessed from 1 July 2009 to 30 June 2010. Lucerne had the highest total annual yield of 12880 kg DM/ha/y, followed by CF/Sub at 9460 kg DM/ha/y. All other pastures produced 6600 to 8360 kg DM/ha/y. Soil water measurements showed that all pastures had a plant available water content of 316 ± 24 mm. All pastures used 655 ± 22 mm/y of water. Lucerne had the highest annual water use efficiency (WUE) of 23 kg DM/ha/mm/y. The annual WUE of CF/Sub was 16 kg DM/ha/y and the lowest was 11 kg DM/ha/mm/y from RG/Wc and CF/Wc pastures. The CF/Sub and CF/Bal pastures had the highest total annual legume (sown and volunteer white clover) content of all grass based pastures at 18%. Because of yield differences CF/Sub had the highest annual nitrogen yield of 216 kg N/ha of the grass based pastures although it was less than half that from lucerne (462 kg N/ha). RG/Wc pastures had the highest proportion of weeds which represented 62% of total dry matter in Year 8 (2009/10). For dryland farmers spring is the crucial period for production before growth slows when soil water becomes limiting. For the spring period (1/7/2009 to 29/1/2010) lucerne produced 11120 kg DM/ha, which was the highest yield of all pastures follwed by CF/Sub (1/7/2009 to 16/11/2009) which produced 6020 kg DM/ha. When dry matter production was regressed against thermal time, CF/Sub pastures grew at 5.7 ± 0.08 kg DM/° Cd between 1/7/2009 and 16/11/2009 compared with lucerne which grew at 4.2 ± 0.46 kg DM/° Cd between 1/7/2009 and 29/1/2010 which was an extra 74 days of linear spring growth. WUE during the initial period for lucerne (1/7/09 to 29/1/10) was 22.3 ± 0.5 kg DM/ha/mm (R2=1.00). The CF/Sub pastures produced 20 ± 1.5 kg DM/ha/mm (R2=0.98) from 1/7/09 to 16/11/09 from 306 mm of water used. The lowest WUE was 11.7 ± 0.21 kg DM/ha/mm by CF/Wc pastures. Insect pests within CF/Sub, CF/Wc, RG/Wc and lucerne pastures were quantified in winter. Pests included Argentine stem weevil, clover root weevil, Sitona weevil and grass grub. Argentine stem weevil populations were highest in CF/Sub pastures (63 weevils m⁻²) whereas clover root weevil populations were highest in CF/Wc (5 weevils m⁻²). Grass grub was present in all grass based pastures and was the highest in RG/Wc with 208 grubs m⁻². Present pest levels were not at damaging thresholds. Based on the results from this research, dryland farmers should maximize the use and potential of lucerne on farm and sow cocksfoot as the main grass species due to persistence and insect tolerance compared with perennial ryegrass. Subterranean clover should be the main companion legume within cocksfoot pastures. Increased legume content ensures scarce water is used more efficiently because N deficiency in the grass is alleviated.

This study consisted of two sheep grazed dryland pasture experiments. Experiment l compared sheep production from 3-year-old cocksfoot based pastures grown in combination with white, Caucasian, subterranean or balansa clover with a ryegrass-white clover pasture and a pure lucerne forage. Sheep liveweight gain per head from each pasture treatment and the pure lucerne stand was recorded in the 2006/07 and 2007/08 seasons. The cocksfoot-subterranean clover pasture provided equal (381 kg LW/ha in 2006) or higher (476 kg LW/ha in 2007) animal production in spring and gave the highest total animal production (646 kg LW/ha) averaged across years of the five grass based pastures. However, total annual liveweight production from lucerne was higher than any grass based pasture mainly due to superior animal production during summer when lucerne provided 42-85% higher animal production than any of the grass based pastures. In Experiment 2, the effect of stocking rate (8.3 (low) and 13.9 (high) ewes + twin lambs/ha) and time of closing in spring on lamb liveweight gain, pasture production and subterranean clover seedling populations was monitored over 2 years for a dryland cocksfoot-subterranean clover and ryegrass-subterranean clover pasture in Canterbury. In both years, twin lambs grew faster (g/head/d) in spring at low (327; 385) than high (253; 285) stocking rate but total liveweight gain/ha (kg/ha/d) was greater at high (7.26; 7.91) than low (5.43; 6.38) stocking rate. Ewes also gained 0.5 and 1.5 kg/head at the low stocking rate in 2006 and 2007 respectively but lost 0.2 kg/head in 2006 and gained 0.3 kg/head at high stocking rate in 2007. Mean subterranean clover seedling populations (per m²) measured in autumn after grazing treatments in the first spring were similar at both low (2850) and high (2500) stocking rate but declined with later closing dates in spring (3850, 2950, 2100 and 1700 at 2, 4, 6, 8 weeks after first visible flower). Seedling populations measured in autumn after grazing treatments in the second spring were also unaffected by stocking rate (low 1290, high 1190) but declined with later closing dates in spring (1470, 1320 and 940 at 3, 5 and 8 weeks after first flowering, respectively). The effect of stocking rate and closing dates in spring on pasture and clover production in the following autumn was similar to the effects on seedling numbers in both years. However, clover production in the following spring was unaffected by stocking rate or closing date in the previous year at the relatively high seedling populations generated by the treatments. This was presumably due to runner growth compensating for lower plant populations in pastures that were closed later in spring. Subterranean clover runner growth in spring may not compensate in a similar manner if seedling numbers in autumn fall below 500/m². Mean annual dry matter production from cocksfoot and ryegrass pastures grown with and without annual clovers pasture production ranged from 6.4 to 12.4 t DM/ha/y but stocking rate (8.3 vs. 13.9 ewes/ha) during spring did not affect annual pasture production. Pastures overdrilled with annual clovers yielded 23-45% more dry matter production than pastures grown without annual clovers. The study confirms the important role of subterranean clover in improving pasture production and liveweight gains of sheep in dryland cocksfoot and ryegrass pastures. Lowering stocking rate from 13.9 to 8.3 ewes/ha was a less effective method of increasing seed production of subterranean clover in dryland pastures although it did lead to increased liveweight gain per head.

The annual and seasonal water use efficiency of six pasture combinations were calculated from the ‘MaxClover’ Grazing Experiment at Lincoln University. Pastures have been established for six years and are grazed by best management practices for each combination. Measurements for this study are from individual plots of four replicates of ryegrass (RG)/white clover (Wc), cocksfoot (CF)/Wc; CF/balansa (Bal) clover; CF/Caucasian (Cc) clover; CF/subterranean (Sub) clover or lucerne. Water extraction measurements showed soils for all dryland pastures had a similar plant available water content of 280±19.8 mm. Dry matter measurements of yield, botanical composition and herbage quality were assessed from 1 July 2008 until 30 June 2009. Lucerne had the highest annual yield of 14260 kg DM/ha/y followed by the CF/Sub at 9390 kg DM/ha/y and the other grass based pastures at ≤ 6900 kg DM/ha/y. All pastures used about 670±24.4 mm/y of water for growth. Lucerne had the highest annual water use efficiency (WUE) of 21 kg DM/ha/mm/y of water used (total yield/total WU). The WUE of CF/Sub was the second highest at 15 kg DM/ha/mm/y, and the lowest was CF/Wc at 9 kg DM/ha/mm/y. The CF/Sub pastures had the highest total legume content of all grass based pastures at 21% and as a consequence had the highest annual nitrogen yield of 190 kg N/ha. This was lower than the monoculture of lucerne (470 kg N/ha). Ryegrass/white clover had the highest total weed component in all pastures of 61%. For dryland farmers spring is vital for animal production when soil temperatures are rising and moisture levels are high. The water use efficiency at this time is important to maximize pasture production. In spring lucerne produced 8730 kg DM/ha, which was the highest dry matter yield of all pastures. The CF/Sub produced the second highest yield of 6100 kg/DM/ha. When calculated against thermal time, CF/Sub grew 5.9 kg DM/ºCd compared with lucerne at 4.9 kg DM/ºCd. The higher DM yield from lucerne was from an extra 400 ºCd of growth. The highest seasonal WUE of all pastures occurred in the spring growing period. Linear regressions forced through the origin, showed lucerne (1/7/08-4/12/08) had a WUE of 30 kg DM/ha/mm (R2=0.98). Of the grass based pastures, CF/Sub produced 18 kg DM/ha/mm (R2=0.98) from 1/7 to 10/11/08 from 270 mm of water used. The lowest spring WUE was 13.5 kg DM/ha/mm by CF/Bal pastures which was comparable to the 14.3±1.42 kg DM/ha/mm WUE of CF/Wc, CF/Cc and RG/Wc pastures. During the spring, CF/Sub clover had the highest spring legume component of the grass based pastures at 42% and produced 120 kg N/ha. This was lower than the 288 kg N/ha from the monoculture of lucerne. Sub clover was the most successful clover which persisted with the cocksfoot. Based on the results from this study dryland farmers should be encouraged to maximize the potential of lucerne on farm, use cocksfoot as the main grass species for persistence, rather than perennial ryegrass, and use subterranean clover as the main legume species in cocksfoot based pastures. By increasing the proportion of legume grown the water use efficiency of a pasture can be improved. When pastures are nitrogen deficient the use of inorganic nitrogen may also improve pasture yields particularly in spring.

This research examined seasonal patterns of growth and development of lucerne crops grown with different levels of crown and taproot reserves in a cool temperate environment. The approach required derivation of explanatory mathematical relationships between crop physiological processes and the main environmental variables of temperature, photoperiod and incoming radiation. To create crops with contrasting levels of perennial reserves (i.e. carbon and nitrogen stored in crowns and taproots), four defoliation treatments were applied to an established 'Kaituna' lucerne crop at Lincoln University, Canterbury, New Zealand in the 2002/03 and 2003/04 growth seasons. Treatments consisted of a factorial combination of a (i) 28 day (short, S) or (ii) 42 days (long, L) regrowth cycle during (i) spring/mid-summer and/or (ii) mid-summer/autumn. The treatments had acronyms of LL, LS, SL or SS, in which each letter refers to the frequency of defoliation in the first and second period of the season, respectively. Regardless of defoliation treatment, perennial dry matter (DMper) differed seasonally. In LL crops, DMper increased from 5 t/ha in autumn. Frequent defoliations (SS) reduced both the DMper by 20-30% and the concentration and amount of soluble sugars, starch and nitrogen in taproots. As a result, the annual shoot yield ranged from 23 t/ha (LL) to 14 t/ha (SS) in 2003/04. Most of this difference was explained by changes in the weight of each individual shoot. Despite shoot yield differences, plant population declined at similar exponential rates in all crops from 140 plants/m² in August 2002 to 60 plants/m² in October 2004. Final shoot populations were conservative at ~780 shoots/m² in all crops. Physiologically, yield differences were mostly explained (R²=0.84) by the accumulated intercepted photosynthetic active radiation (ΣPARi). The ΣPARi was limited in frequently defoliated crops because these crops had lower leaf area index (LAI) than LL crops but similar canopy architecture, with an extinction coefficient for diffuse light (kd) of 0.81. Differences in LAI were a direct effect of harvesting before full canopy cover (i.e. critical LAI of 3.6) in short regrowth cycles. In addition, leaf area expansion rates (LAER) in spring were reduced from 0.016 (LL) to 0.011 m²/m²/°Cd (SS) as taproot DM declined from 3.0 to 1.5 t/ha, respectively. The reduction in LAER was a result of smaller leaves after the 5th main stem-node and a reduction in the leaf area of branches. At the same time the phyllochron was conservative at 34°Cd/leaf (Tb of 5°C) when photoperiod was greater than 12.5 h but increased to 40°Cd (LL and SL) and 60°Cd (LS and SS) at lower photoperiods. All crops had a similar maximum rate of branching of 6.8 leaves/main-stem node and leaf senescence rates at 0.48 leaves/main-stem node. Shoot radiation use efficiency (RUEshoot) was 1.6 g DM/MJ PARi for pooled annual yield. However the seasonal RUEshoot was inconsistent being 1.8 g DM/MJ PARi in spring/summer and only 1.0-1.2 g DM/MJ PARi in autumn for LL crops. Frequent defoliations decreased RUEshoot by 20% during summer. The average RUE for total dry matter (RUEtotal) was 2.2 g DM/MJ PARi but this was affected by defoliation treatments and temperature. Measurements of RUE from 23 regrowth cycles were then used to validate a previously proposed temperature framework for RUE. That framework was used to estimate the RUEtotal and fractional partitioning of DM to perennial organs (Pper). In SS crops, RUEtotal was 50% of the values of LL crops in four of eight analysed regrowth cycles. This was consistent with the measured reduction of 20% in the saturated photosynthetic rate (Pmax) of SS crops in the summer/autumn of 2002/03 season. Photosynthetic rates at 1000 µmol photon/m²/s (Pn1000) were also ~20% lower during the first 150°Cd of regrowth in SS crops in comparison with LL crops. Differences in Pn1000 were mostly explained by the specific leaf nitrogen and chlorophyll content of these leaves. The estimated Pper increased from

Lucerne (Medicago sativa L.) was found to be a more productive dryland forage option than chicory (Cichorium intybus L.) or red clover (Trifolium pratense L.). This was concluded from superior annual dryland yields of 20 t DM/ha from lucerne compared with 14-16 t DM/ha for chicory and red clover. This yield advantage was achieved by higher growth rates during both cool spring/autumn periods and dry summer periods. Lucerne was also the most persistent species maintaining a botanical composition of 94% six seasons after establishment, compared with 65% for chicory and 0% for red clover. All three species had similar herbage quality (25% crude protein, 11.5 MJ ME/kg DM) and grazing stock consumed 30% more protein and energy from lucerne than chicory or red clover crops. The superior lucerne production during dry periods was due to increased water extraction up to 2.8 m depth, compared with ~1.9 m for chicory and red clover. All three crops displayed a top-down perennial water extraction pattern with an extraction front velocity of ~15 mm/day. Depletion of available water capacity in each layer of the soil profile was exponential following the arrival of the extraction front. A detailed examination of lucerne physiology was conducted to understand seasonal variation, and the effects of water shortages on forage yield. Total DM production under non-water and non-temperature limiting conditions was related to total intercepted radiation. The total radiation use efficiency (RUE) was found to be 1.6 g/MJ. However, there was a seasonal change in DM partitioning between shoot and perennial organs (roots and crowns) and its influence on forage yield was quantified by converting total RUE to shoot RUE. The shoot RUE was 1.3 g/MJ in September, gradually decreased to a constant 1.0 g/MJ from mid-December late-January and then abruptly decreased to 0.6 g/MJ in March/April. Temperature also influenced shoot production and this was quantified by multiplying RUE by a linear factor that declined from unity at a mean regrowth cycle air temperature of 18°C to zero at 0°C. Seasonal changes in radiation interception were quantified by studying the influence of temperature and photoperiod (Pp) on the components of leaf area index (LAI) expansion. Specifically, main-stem node appearance was linear in response to Tt and the phyllochron was 37±7°Cd for from August-January. However, phyllochron increased to 60 °Cd when the Pp on the day 150 °Cd before the first node decreased to 16 h (24 January). Continued decrease in Pp gave a 5.6 °Cd/h Pp reduction in phyllochron returning, it to 37±7 °Cd at a Pp of 13.5 h (15 March). There was a poor relationship between main-stem node appearance and LAI expansion, suggesting branching and leaf expansion have different seasonal responses to environment. Water shortages were quantified by crop transpiration (ET) relative to the crops ET demand. Crop ET was calculated from water balance by removing evaporation losses from the soil and outer canopy. Crop ET demand (EPT) was calculated from Penman evapotranspiration potential (EP) multiplied by crop cover and a calibration coefficient (0.86), determined by regressing the ET of irrigated crops against EP. The RUE and LAI of dryland crops was expressed as a fraction of irrigated crops (fD/I) to quantify the effects of water stress. The LAI expansion of lucerne was the most sensitive process with fD/I of 1.0 at an ETIEPT of 0.97 decreasing to 0.1 at an ET/EPT of 0.22. There was a 1: 1 relationship between the fD/I of RUE and ET/EPT. It is concluded that the improved understanding of lucerne environmental responses presented in this thesis must be considered when examining yield variability of lucerne.

Silvopastoral systems comprise part of the continued expansion of conifer plantings on grassland in New Zealand. Previous studies of afforestation and agroforestry in New Zealand have not adequately addressed the effects of interactions between over and understorey species on soil phosphorus (P) dynamics and availability. Accordingly, the main objective of this study was to investigate soil P dynamics and availability under a temperate conifer-based silvopastoral system in New Zealand. This involved field and glasshouse experiments which examined the effects of combinations of Pinus radiata, lucerne (Medicago sativa L.) and ryegrass (Lolium perenne L.) on soil P. Thus spatial and temporal aspects were examined under field conditions, while the influence of soil properties was investigated under controlled conditions. The main silvopastoral system examined in this study was established at Lincoln University in 1990 on soil that had previously been managed under intensive arable cropping. The investigation of spatial variation found that the more labile inorganic P fractions increased while organic P fractions decreased, under trees compared to forages, close to the ripline in the tree row. These changes tentatively indicated that trees had enhanced mineralisation of soil organic P. However, organic P fractions appeared to be greater in the ripline under trees grown without understorey vegetation. This apparent contradiction was attributed to the competing effects of enhanced mineralisation of organic P by trees, compared to forages, and increased accumulation of organic P under trees and forages as a consequence of conversion to perennial vegetation from intensive cropping. Although soil samples were collected to a depth of 60 cm, the major changes in soil P occurred in the surface 20 cm. Changes in soil P over the first 10 years of the Lincoln silvopastoral system were determined on samples taken in 1991, 1994, 1996 and 2000. The findings of other workers that plant available inorganic P was enhanced and organic P reduced under conifers were not confirmed in this study. This was mainly attributed to the fact that soil organic carbon and P increased with time at the Lincoln site following conversion to perennial vegetation from intensive cropping. A study of the rhizosphere was carried out on the Lincoln agroforestry trial over a one year period. Lower levels of labile organic P, and higher levels of labile inorganic P and phosphatise enzyme activity were determined in tree and forage rhizosphere soils in spring compared with autumn. This data confirmed that overall rates of soil organic P mineralisation were greater in spring. In addition, concentrations of labile inorganic P were consistently greater in tree rhizosphere soil compared to the companion understorey. Lucerne-tree rhizosphere soil and lucerne rhizosphere soil contained the lowest concentrations of total inorganic and organic P compared with ryegrass-tree and ryegrass rhizosphere soils. This was partly attributed to higher levels of phosphatase enzyme activity in the lucerne rhizosphere soils. It became apparent from the spatial and temporal studies that initial soil properties would likely influence soil P changes. Accordingly a detailed glasshouse experiment was carried out to examine the short-term effects of different soil properties on combinations of Pinus radiata, lucerne and ryegrass. This included an investigation of the effect of tree ectomycorrhizal hyphae on soil P. Although P uptake was greatest under radiata pine, trees tended to deplete inorganic P to a lesser extent than the forages. Isotopic exchange kinetics analysis revealed that trees and lucerne combined caused a greater decline in the most recalcitrant forms of soil inorganic P. This enhanced depletion of recalcitrant inorganic P under trees and lucerne may have been related to increased N availability. There was also evidence of significant redistribution of soil P from less labile to more labile forms. Where levels of P and carbon were very low, trees and lucerne were able to deplete most P forms, and when lucerne was combined with trees, depletion of recalcitrant inorganic P increased. Results indicated that in low P soils redistributed P was mainly taken up by plants, whereas concentrations of available inorganic P increased in high P soils. In high P soils labile organic P was not depleted, suggesting that redistribution of moderately available and recalcitrant P to available inorganic and organic forms had occurred. An important aspect of the effect of Pinus radiata on soil P was the ability of ectomycorrhizal hyphae to substantially deplete most P fractions, thus confirming the important role that ectomycorrhizal fungi play in determining P uptake by trees. The findings of this study confirm that changes in soil P forms are strongly influenced by interactions between plant species and soil properties which, in turn, are influenced by land use and management.

The general goal of this research was to understand the agronomic and physiological changes of a lucerne crop in distinct physical radiation environments and to verify the potential of lucerne to grow under shaded conditions. To achieve this, the research was conducted in four main steps: (i) firstly, experimental data collection in the field using two artificial shade materials (shade cloth and wooden slats) under inigated and non-irrigated conditions; (ii) a second experiment with data collection in a typical temperate dryland agroforestry area under non-irrigated conditions; (iii) generation of a light interception sub-model suitable for shaded crops and (iv) a linkage between the light interception sub-model and a canopy photosynthesis model for agroforestry use. In experiments 1 and 2, lucerne crop was exposed to 6 different light regimes: full sunlight (FS), shade cloth (FS+CL), wooden slats (FS+SL), trees (T), trees+cloth (T +CL) and trees+slats (T+SL). The FS+SL structure produced a physical radiation environment (radiation transmission, radiation periodicity and spectral composition) that was similar to that observed in the agroforestry site (f). The mean annual photosynthetic photon flux density (PPFD) was 41 % under the FS+CL, 44% under FS+SL and 48% under T compared with FS in clear sky conditions. Plants were exposed to an intermittent (sun/shade) regime under both FS+SL and T, whereas under FS+CL the shaded light regime was continuous. The red to far-red (RIFR) ratio measured during the shade period under the slats was 0.74 and under the trees was 0.64. However, R/FR ratio increased to 1.26 and 1.23 during the illuminated period under FS+SL and T, respectively, and these were equivalent to the ratio of 1.28 observed under the FS+CL and 1.31 in FS. The radiation use efficiency (RUE) of shoots increased under the 5 shaded treatments compared with full sunlight. The pattern of radiation interception was unchanged by radiation flux, periodicity and spectral composition and all treatments had a mean extinction coefficient of 0.82. However, the magnitude of the decrease in canopy growth was less than those in PPFD transmissivity. The mean lucerne annual dry matter (DM) yield was 17.5 t ha⁻¹ in FS and 10 t ha⁻¹ under the FS+CL, FS+SL and T regimes. This declined to 3.4 t DM ha⁻¹ under T+CL (22% PPFD transmissvity) and 4.1 t DM ha⁻¹ under T+SL (23% transmissivity). A similar pattern of response was observed for leaf net photosynthesis (Pn) rates under the shade treatments compared with full sun. In addition, spectral changes observed under the trees and slats affected plant motphology by increasing the number of long stems, stem height and internode length compared with full sunlight. Thus, there were two main explanations for the increase in RUE under shade compared with full sun: (i) preferential partition of assimilates to shoot rather than root growth and/or (ii) leaves under shade were still operating at an efficient part of the photosynthetic light curve. The changes proposed for the canopy Pn model were appropriate to simulate the radiation environment of an agroforestry system. However, the model underestimated DM yields under the continuous and intermittent shade regimes. These were considered to be mainly associated with plant factors, such as overestimation in maintenance respiration and partitioning between shoots and roots in shade and the intermittency light effect on leaf Pn rates. Further investigation in these topics must be addressed to accurately predict crop yield in agroforestry areas. Overall, the lucerne crop responded typically as a sun-adapted plant under shade. It was concluded that lucerne yield potential to grow under intermediate shade was superior to most of C3 pastures previously promoted in the literature.

En este trabajo se describe una plagadeSminthurus viridis L. sobre la alfalfa de los campos de la ribera del rio Arga en Navarra. La plaga apareció en el otoño de 1997. Junto a S. viridis apareció, en grandes cantidades, Entomobrya schuetti Stach, 1922. Esta última especie se encontraba alimentándose de los hongos que crecían en las heridas producidas por S. viridis en las hojas de la alfalfa. Se contaron los individuos de ambas especies en un 10% de la muestra. La proporción de ambas es del 26% (S. viridis) y 74% (E. schoetti) respectivamente. Se ha estimado la población en uno de los campos estudiados en unos 2.000 ejemplares por planta de alfalfa.

Two experiments were conducted with sheep to determine the nutritional value of 'Lathco' flatpea (Lathyrus sylvestris) hay. In experiment 1, a finishing trial was conducted with 50 group fed crossbred lambs (25 ewes and 25 wethers) fed diets consisting of 70% chopped forage and 30% ground corn grain for 70 d. Five pens of five wethers and five pens of five ewes were selffed diets in which the forage consisted of 100:0, 75:25, 50:50, 25:75, and 0:100 proportions of alfalfa (Medicago sativa L.) and flatpea hay. Feed efficiencies decreased linearly (P

In order to ensure that strains of rhizobia that are efficient nitrogen fixers nodulate cultivated legumes in the field, it is important to be able to identify those that can compete effectively with indigenous strains in the fonnation of nodules. To this end, experiments have been conducted to investigate whether the timing of infection of legume plants by rhizobia is related to the ability of the strains to compete for nodule occupancy. The periods that 16 strains of R. meliloti in stationary growth phase took to infect lucerne (Medicago sativa, L.) plants were evaluated by estimating the time interval between inoculation of plants and initiation of the uppermost tap root nodule. Subsequently, nodulation competitive ability of each strain was evaluated in 29 paired-strain experiments. The results from each pair were compared to the periods to infection by the constituent strains. Strains which infected plants significantly faster (p < 0.05) were more competitive in 8 of 11 pairs; the same strain was more competitive than could be predicted on the basis of timing of infection in the remaining 3 of the 11 pairs. However, when the strains showed no difference in periods to infection, they had similar nodulation competitiveness in only 2 of 18 pairs. The effects of inoculating plants with exponential-phase cells, and of incubating strains in rhizosphere or in luteolin solution before inoculation of plants, were investigated. Exponential phase cells infected plants more quickly than stationary phase cells, but incubation of cells in luteolin solution did not affect timing of infection. Cells that had been incubated in rhizosphere, i.e. in plant nutrient solution containing growing lucerne roots, were slower at infecting plants than cells that had been incubated in plant nutrient solution alone. In the treatments where different periods to infection were observed in these experiments, strains which were quicker to infect plants in single-strain inoculation experiments were also more competitive in paired-strain experiments. Antagonism between the strains was evaluated to find out whether it was a factor that could influence competitive ability. The results showed that antagonism was of minor importance compared with timing of infection as strains showing a higher degree of antagonism were more competitive in nodule formation in only 4 of 15 pairs, and all the four pairs were in the category (consisting of 16 pairs) that showed no differences in timing of infection between component strains of each pair. Prior inoculation of plants with one strain reduced the number of nodules formed by strains inoculated subsequently by 39 to 68 %, but the strains inoculated subsequently did not differ in competitive ability from those without prior inoculation. These results suggest that strains of rhizobia which infect plants more rapidly are often more competitive because they form the first nodules, which induce the autoregulatory system of the plant to limit further nodulation, thereby reducing the opportunities of slower strains to form nodules. The results of this investigation indicate that it is possible to rank strains of rhizobia for nodulation competitiveness by screening them for time taken to infect plants, but these results need to be tested in the soil environment.

The effect of lime on the growth and nodulation of lucerne (Medicago sativa L. cv. AS13R) was studied on a Templeton silt loam at Lincoln. Measurements were made of the shoot dry matter yield, root dry weight, and the number of nodules. Shoot nitrogen percentage at the second harvest was analysed. Increasing rates of lime gave increased shoot dry matter yields generally, but the size of the response depended on the time of sampling during growth.Yield increases from 0 t ha⁻¹ to 8 t ha⁻¹ lime (pH 5.3 - 5.4 to 6.1 - 6.2) were 35 - 37% over the first two harvests. Lime did not significantly increase root dry weight to a depth of 20 cm. Lime increased the total number of nodules at all samples. The number of small (< 2 mm) nodules increased significantly with lime at all samples. During the first harvest, lime gave a response with the number of medium (2 - 4 mm) and large (> 4 mm} nodules, but there was no response to lime during the second harvest. The total number of nodules increased with time at all lime rates . Lime did not affect shoot N % at the second harvest, but gave a highly significant increase in N yield.

In the second year of a two year grazing management trial on Wairau Lucerne, plots were grazed under a 3 or 6 weekly rotation. Half of each plot was nard grazed in June to control over wintering aphids. In the spring/autumn an aphicide treatment was applied. Plant roots excavated in the autumn, were used for determination of plant populations, root dry weights and analysis for non-structural carbohydrate root reserves. Cool season grazing management reduced spring aphid populations from 37 to 3/stem and showed no significant difference in dry matter yield. Three weekly spelling, although reducing aphid numbers, halved plant populations, root weights, and markedly reduced dry matter yield. Associated with this decline, was the dramatic increase in weed populations, notably Scotch thistle. Weeds contributed 60% to the total yield of 10,400 kg/ha under 3 weekly spelling. However, the shading effect of the lucerne canopy and harsh conditions following grazing, reduced weed populations and prevented thistle establishment throughout the growing season, under 6 weekly spelling, resulting in only 23% of 18,800 kg/ha total yield being weeds. Aphid control by spraying throughout the growing season increased lucerne yield, but not significantly. Spraying of lucerne in the autumn reduced aphid populations from 21/stem to 1/stem under 6 weekly spelling and resulted in high plant root reserves, with no significant effect on top yield. Only under spraying, was there a significant difference in root reserves between 3 and 6 weekly spelling. The results show that cheap methods of controlling aphids in lucerne can be developed using grazing management.

A study was conducted on pure WL320 luceme, luceme/'Grasslands Matua' prairie grass and lucerne/'Grasslands Maru' phalaris mixtures at Lincoln University, between March 20 and April 7, 1991. The 0.042 ha plots were sown in early November 1986 into a Templeton fine sandy loam soil. Irrigation immediately before the study started attempted to bring plots to field capacity. The results showed no significant differences in dry matter yields, total water use and water use efficiency between the three pasture types. Although not significant dry matter production was higher in pure lucerne (58 kgDMha⁻¹d⁻¹) than lucerne/prairie grass (51 kgDMha⁻¹d⁻¹) and lucerne/phalaris (45 kgDMha⁻¹d⁻¹). Water use efficiency was 14.1, 12.3 and 10.2 kgDMha⁻¹mm⁻¹H₂O for pure luceme, lucerne/prairies grass and lucerne/phalaris respectively. Deep soils produced more dry matter (56 vs 47 kgDMha⁻¹d⁻¹) and had a higher water use efficiency (12.7 vs 11.7 kgDMha⁻¹mm⁻¹H₂O) than shallow soils although these differences were not significant.

Above ground biomass production, crown architecture and canopy development were studied in seedling origin (seedling trees) and a tissue culture origin clone (clonal trees) of Pinus radiate trees, growing with and without pasture understorey (Medicago sativa). The objective was to improve our understanding of above ground processes influencing tree growth in an agroforestry situation. This study was carried out at the Lincoln University Agroforestry Experiment located in Lincoln (43° 38' S, 172° 30’ E; 11 m a.s.l.), New Zealand, where the climate is temperate with uniformly distributed annual rainfall of about 660 mm and the soil is a free draining silt loam. Although, the clonal trees were 0.5 m taller and had greater stem basal area than the seedling trees at age 3, there were no differences in accumulated total above ground biomass and biomass increment. However, due to differences in allocation of biomass to crown components, the clonal trees had a higher harvest index and a higher foliage efficiency (FE: biomass production per unit foliage area). The crown architectural study found that the clonal trees had a higher branch basal area and more medium size branches than the seedling trees. The branch basal area to foliage area relationship indicated that the clonal trees had less foliage area per unit branch basal area than the seedling trees. They also had longer internodes and a higher crown shape ratio (crown height/crown width) than the seedling trees. These differences in crown architecture and foliage distribution affected the FE of the different planting materials. When trees were grown with the lucerne understorey, the accumulated above ground biomass was decreased by 67% compared to the no-understorey treatment. But the reduction in basal area and stem wood mass were only 42 and 50%, respectively. Allocation of above ground biomass into crown components was also affected by planting materials. Wood fraction was higher and branch fraction was lower in clonal trees than seedling trees resulting in a higher harvest index. Branch basal area and foliage area relationship indicated less foliage area per unit branch basal area, leading to improved FE. These changes were greater in the seedling than the clonal trees. These changes in biomass and crown architecture were extended to look at the interception of solar radiation, the efficiency of carbon assimilation and productivity after the live crown pruning of trees. Growth losses due to pruning were higher in the clonal than in the seedling trees when they were grown in the no-understorey treatment. But due to the hindrance of tree growth by lucerne competition, growth losses due to pruning were smaller than where there was no-understorey. New foliage growth was related to water stress of the trees, but in pruned trees the effects of water stress were decreased. However, there was no effects of pruning on the temporal growth pattern of foliage and branches, specific leaf area and needle density, nor, on above-ground biomass allocation to foliage, branches and stem wood. Pruning also resulted in an increased FE. This improved FE was due to increased photosynthesis rate (thus light conversion efficiency) and increased light interception per unit foliage area. Photosynthesis and light conversion efficiency was increased more than 30% and 35%, respectively, in pruned trees compared to the unpruned trees. Light intercepted per unit foliage area was about 35% higher in pruned trees, but n decreased with increasing foliage area after pruning. More detailed light interception studies are needed in agroforestry systems to determine the optimum pruning strategy, taking into account of impact on tree and understorey growth and wood quality. Forest managers usually manipulate stem and branch growth by altering planting density. However, this study found that understorey competition can be used in agroforestry situations to achieve desirable tree characteristics such as reduced branch size and improved harvest index.

Mixtures of WL 320 lucerne (Medicago sativa L.) with Matua prairie grass (Bromus willdenowii Kunth) or Maru phalaris (Phalaris aquatica L.) were compared with pure lucerne over a one and half year period. Two grazing duration treatments were compared: short grazing duration (three to seven days) and long grazing duration (six to 14 days) to give a 3 x 2 factorial randomized block design laid down in four replications with plots of 0.042 ha. Pasture seeds were drilled into Templeton silt loam in 15 cm rows on November 7, 1986 at rates of 32, 8 and 7.5 kg/ha for Matua prairie grass, Maru phalaris and lucerne. Grasses were sown across the lucerne rows. Limited irrigation (25 mm + 40 mm) was applied 43 days and 66 days after sowing to supplement the lower than average summer rainfall. No irrigation was used after grazing commenced. The 144 day establishment phase ranged from the sowing date (November 7, 1986) to the commencement of the first grazing duration treatment (April 1, 1987). All plots were grazed in common for seven days from January 25 to January 31, 1987. During the grazing duration treatment phase (from April 1, 1987 to January 14, 1988) four grazings with sheep were monitored to investigate the possibility of manipulating the lucerne/grass balance by using short or long grazing durations. Because grass tillers recover more rapidly after defoliation compared with lucerne shoots it was hypothesized that long grazing durations would favour lucerne and short durations would result in grass dominance. The lucerne/prairie grass mixture produced 11,150 kg DM/ha during the 144 day establishment phase compared with 6,480 DM /ha for lucerne/phalaris and 6740 kg DM/ha for pure lucerne. The contribution of lucerne to total yield of lucerne/prairie grass and lucerne/phalaris mixtures was 16% and 53% respectively. Lucerne/grass mixtures provided better ground coverage and depressed weed invasion compared with pure lucerne. There was no significant difference in total dry matter yield between the three pasture types during the grazing treatment phase (April 1, 1987 - January 14, 1988), the yields being 13,330, 12,710 and 13,900 kg DM/ha for lucerne/prairie grass, lucerne/phalaris and pure lucerne respectively. However, the total lucerne dry matter yield in lucerne/prairie grass (4,130 kg DM/ha) was half of that in lucerne/phalaris (8,450 kg DM/ha), but total prairie grass yield (9,200 kg DM/ha) was over twice that of the phalaris (4,260 kg DM/ha). All pasture types produced the highest total single harvest yields in the October harvest and lowest at the July harvest, with the herbage yield being 5,450, 4,760 and 4,800 kg DM/ha for lucerne/prairie grass, lucerne/phalaris and pure lucerne at the October harvest, and 1,870, 1,700 and 1,300 kg DM/ha at the July harvest. Lucerne/prairie grass and lucerne/phalaris mixtures produced 45% and 31% more dry matter than pure lucerne in the first winter. Both lucerne/prairie grass and lucerne/phalaris were dominated by lucerne (64% and 96% lucerne in lucerne/prairie grass and lucerne/phalaris) in January/88 harvest, but by grass (95% and 78%) at July harvest. Long grazing duration significantly depressed total grass yield in the lucerne/prairie grass by 2500 kg DM/ha, but by only 900 kg DM/ha in lucerne/phalaris mixture. At both the October and November harvests long grazing duration depressed prairie grass yield by 1150 and 820 kg DM/ha, and phalaris yield by 540 and 250 kg DM/ha respectively. Long grazing duration increased lucerne dry matter yield in lucerne/prairie grass at November cut. Generally, long grazing duration increased lucerne shoot numbers, shoot size and shoot height, but decreased grass tiller numbers in lucerne/grass mixtures, particularly with lucerne/prairie grass. There was no significant effect of grazing duration on grass tiller size and tiller length. However, the total DM yield from long grazing duration was 13% less than that from short duration (12,400 kg DM/ha vs 14,250 kg DM/ha for long and short durations) because of the confounding effect of shorter regrowth periods with less mature pastures after long grazing duration. Urine nitrogen stimulated grass growth and increased grass tiller numbers and yield, which in tum depressed lucerne growth and decreased lucerne shoot numbers and yield in lucerne/grass mixtures. Prairie grass was most responsive to urine because the lower lucerne proportion in the mixture resulted in lower nitrogen inputs to the pasture and lower nitrogen content of herbage (2.5%) compared with 3.1% for lucerne/phalaris mixture and 3.5% for pure lucerne. These results suggested that long grazing durations of lucerne/grass mixtures may encourage lucerne and that short duration will encourage grasses.

Fine root length density and coarse root biomass and architecture of tissue cultured clonal and seedling planting materials of Pinus radiata D.Don was estimated in an agroforestry system with understorey treatments (no understorey, Medicago sativa, and Lolium perenne + white clover) at the Lincoln University agroforestry experiment site, New Zealand. Seasonal growth patterns of fine and coarse roots, branches and needles of radiata pine trees were related to soil moisture and temperature as influenced by planting material, understorey and micro-site conditions. The climate was summer-dry temperate with 660 mm annual rainfall evenly distributed throughout the year. The monthly mean evapo-transpiration was lower than the rainfall in winter, whereas in spring and summer it exceeded rainfall by 250%. Most of the rain-bearing winds come from the south, resulting in the soil on the south side of the tree rows being cooler and wetter compared to the north sunny side. At age 3, clone 3 trees were taller than clone 4 and seedling trees although the basal area was very similar among these planting materials. The fine root length density of clone 3 and clone 4 trees was higher than seedling trees particularly at 0-10 cm depth. The fine root length density was higher in the ripline compared to the north and south side, and was higher on the south side compared to the north side. Fine root distribution was thus a function of planting materials, site preparation and soil moisture distribution. Lucerne had higher rooting density than the ryegrass/clover plants, and so provided greater competition to the trees. Coarse root architecture studies showed that root:shoot ratio, and biomass of root core, vertical and the lateral roots, particularly in the ripline, were higher in clone 3 than in the seedlings. As a result of understorey competition, seedlings had a lower vertical and lateral root biomass in the lucerne treatment compared to the no understorey treatment, whereas in clone 3 root biomass was less affected by lucerne competition. The lateral root biomass was higher in the ripline compared to the sides, and it was higher in the south compared to north. Root biomass growth between 1993 and 1994 was much higher in clone 3 than seedling trees, particularly with the vertical roots; also much higher in the no understorey than with the lucerne treatments. The fractional allocation to the core did not change for clone 3, whereas it was lower for seedlings in 1994 compared to 1993. Topsoil moisture content rather than temperature appeared to be the major factor in control of root growth and development. Higher soil moisture content in the 10 to 30 cm depth was associated with earlier and higher maximum rates of fine root tip initiation than in the 0-10 cm depth, and this resulted in a higher maximum rate of total fine root elongation in the 10-30 cm depth. Similarly, a higher soil moisture content in the no understorey treatment was associated with a higher maximum rate of fine root elongation and coarse root cross sectional area growth, particularly with clone 3 than in the lucerne treatment. The maximum rate of total fine root elongation for clone 3 was 35% higher in the no understorey compared to lucerne; seedlings did not differ with understorey treatments. This illustrated the vigorous root regeneration character of clone 3 compared to seedlings. Also, the maximum rate of fine root elongation occurred earlier in the no understorey than in the lucerne treatments. The effect of lower soil moisture content on the north as opposed to the south side of lucerne and both north and south sides in the case of the no understorey treatments further resulted in a lower maximum rate of total fine root elongation. This indicated that lucerne was very competitive for soil moisture particularly on the north side of the ripline. The growth pattern of the above ground (i.e., fascicle elongation, branch cross sectional area growth) and below ground (white root tips initiation, total fine root elongation and coarse root cross sectional area growth) components followed the sequence: (a) firstly, early spring growth of branches provided room for fascicle emergence and elongation, (b) secondly, fine root initiation occurred in late summer, and cross sectional area of branches and then coarse roots reached their maximum growth, (c) thirdly, fine root elongation occurred in autumn and reached its maximum growth in winter. The result of this study suggest that if the primary aim of tree growers is to produce the maximum amount of timber under an agroforestry system it is necessary to select planting materials with vigorous rooting characteristics in order to support the above ground growth of trees. Planting materials similar to clone 3 seem to be ideal because of their higher fine root length density and coarse root biomass compared to the other planting materials with or without understorey competition. The study of fine root regeneration, and coarse root cross sectional area growth of clone 3 showed a higher maximum growth rate compared to seedlings. These rooting characteristics were linked to the higher above ground growth of clone 3. Furthermore, agroforesters have to select an understorey based on its competitive nature with the tree's root system. In this study, ryegrass/clover as an understorey crop produced less root competition to trees than lucerne.

A review of the literature concerning foliar pathogens of lucerne and the methods used in controlling them has been made with particular reference to fungicides as a means of control. The effect of a systemic fungicide, benomyl, and universal spraying oil on growth and germination of lucerne (Medicago sativa cv Wairau) was studied. Concentrations of 10,000 ppm benomyl and 5% oil were toxic to both lucerne germination and growth. Low concentrations of benomyl (100 ppm) and oil (0.5%) were not toxic as measured by fresh and dry weights or plant lengths. The addition of oil to benomyl did not increase toxicity. The effect of benomyl on growth and germination of Pseudopeziza medicaginis, Phoma medicaginis and Stemphylium botryosum was studied in vitro. Benomyl almost completely inhibited spore germination of Phoma medicaginis at a concentration of 32 ppm and completely inhibited Pseudopeziza medicaginis at 512 ppm. However Stemphylium botryosum spore germination (95%) was not adversely affected by 2000 ppm. Conidia of S. botryosum were found to germinate in the presence of benomyl at 4000 ppm to produce swollen and shortened germ tubes. Mycelial growth of Ph. Medicaginis was completely inhibited at 10 ppm benomyl but mycelia growth of S. botryosum was still recorded at 1000 ppm benomyl. Benomyl (1000 ppm ) and universal spraying oil (1%) as a treatment in glasshouse tests showed that benomyl was not as effective systemically as it was as a protectant or eradicant fungicide. Activity was recorded against Ps. medicaginis S. botryosum. Three field trials were conducted in the 1972/73 season to test the effect of benomyl on: herbage yield, seed yield and seed quality. In two herbage trials plots of lucerne sprayed with, a weekly treatment and an early and late treatment of benomyl (1000 ppm) and oil (1%) yielded 16% and 11% more dry matter than did unsprayed plots (average of two trials). Ps. Medicaginis and Ph. Medicaginis were the major pathogens during the trials and their control reduced defoliation from 30% in untreated to an average of 10.5% in treated plots and increased leaf-stem ratios from 0.65 in untreated to 0.79 and 0.89 in plots sprayed with a weekly and early and late application ( averages of two trials). In a seed yield trial, plots of lucerne were sprayed with a fortnightly treatment and an early late treatment of benomyl (1000 ppm) and universal spraying oil (1% ppm). These plots yielded 13.1% and 44.8% more seed than did unsprayed plots in the same trial defoliation was reduced from 63.9% to 11.8% in the early and late treatment. Dry matter was increased by 20.6% and 22.0% respectively. From treated seed harvested from this trial there was no effect on seed germination of on 1000 seed weights compared to untreated seed. However the incidence of Ph. Medicaginis on seed was reduced from 11.3% in untreated to 7.2% in fortnightly treatments.

Lucerne (Medioago sativa L. 2n=32) is one of the chief forage crop plants and has an increasing importance in the New Zealand farming system. The natural variability in lucerne is considerable. This is mainly because of its polyploid nature and the fact that lucerne is largely an allogamic, self-incompatible species. Brock (1970a and 1972) has pointed out the difficulty regarding the expression and utilization of mutations in cross-fertilizing species and has questioned the usefulness of induced mutations in such species unless it could be clearly established that natural variability has been fully utilized. Since there is no indication of lack of variability in lucerne but rather difficulty in handling it, the use of radiation to induce further variability for plant breeding cannot be justified at present. However, a good case for studying the responses of lucerne to radiation can still be made, partly because little is known about the reactions of this species, and partly in preparation for possible future demand for induced mutations. Biological damage and mutation frequency both increase with increasing levels of radiation. A compromise must therefore be reached between the desired mutation frequency and the degree of biological damage that can be tolerated. Gregory (1968) has shown that the frequency of radiation-induced mutants increases with decreasing magnitude of effect, and this increase in frequency is strongly correlated with the increase in probability that a given change would effect an improvement. Research on the relationship between polyploidy and radiosensitivity carried out so far has shown that, in general, increasing degrees of ploidy are accompanied by increasing radioresistance. With this in mind, an inquiry into the properties of populations of lucerne produced after seed irradiation seemed appropriate and desirable.

The insecticides pirimicarb, fenvalerate, chlorpyrifos and demeton-S-methyl were tested against lucerne aphids Acyrthosiphon kondoiShinji and A. pisum (Harris) and against the natural enemies Austromicromus tasmaniae (Walker) and Coccinella undecimpunctata L. by topical application. LD₅₀ values were determined. Pirimicarb was 1000 – 10000 times more toxic to the aphids than to the predators while the equivalent factors for fenvalerate and demeton-S-methyl were 5 – 1000 and 4 – 100 respectively. Chlorpyrifos showed the same order of toxicity to both predators and aphids. The effect of post-treatment temperatures of 10 and 25°c on the toxicity of pirimicarb, fenvalerate and chlorpyrifos to the adult lacewing A. tasmaniae was investigated. A marked increase in toxicity of fenvalerate was observed at 10°c. A brief investigation of possible sublethal effects of fenvalerate on lacewing fecundity, and of pirimicarb and fenvalerate on the length of the larval developmental stage, revealed no significant effects.

Growth and persistence of C. insidiosum and R. meliloti in soil and rhizosphere, and their interactions with F. oxysporum and V. alboatrum in the infection and nodulation of lucerne plants were investigated. The droplet method, plaque count technique, plant infection technique and the FA technique were used to isolate, identify and count C. insidiosum and R. meliloti in soil. The FA technique was quantitative, sensitive and easy to perform. The plaque count technique was quantitative, its sensitivity depended on the specificity of the bacteriophage used and it required numerous assays for each sampling. The plant infection technique was laborious, time consuming and in general less sensitive than the FA technique and plaque count technique. The droplet method was simple to perform but the lack of selective media for C. insidiosum and R. meliloti rendered it unsuitable for the detection of these bacteria in field soil. Both bacteria could survive for at least 132 d when inoculated into gamma sterilised soil in which persistence wase favoured at soil moisture levels of 58% W.H.C. and 100% W.H.C. and at 9°. In field soil, C. insidiosum persisted for less than 21 d but R. Meliloti was still present after 42 d. C. insidiosum persisted in diseased tissue for more than a year at 52% R.H. but only for 14 d at 100% R.H. C. insidiosum and R. meliloti were stimulated to multiply in the vicinity of roots of lucerne on agar media. In soil, the growth of the two bacteria was stimulated in the rhizosphere when the inoculum levels were 10⁷ cells ml⁻¹ and 10⁴ cells ml⁻¹, but not at 10¹ cells ml⁻¹. Infection of lucerne by C. insidiosum was favoured when the host plants were wounded and at the fourth trifoliate leaf stage than younger stages. The number of plants infected was more at temperature of 21° than 30°, 14° and 10°; at soil moisture levels of 58% W.H.C. and 100% W.H.C. than 30% W.H.C.; at high N and low K conditions. Soil pH in the range of 5.1 to 7.8 did not influence infection. Infection was 90% at inoculum level of 10⁷ cells ml⁻¹ , 50% at 10⁶ cells ml⁻¹ and 0% at 10² cells ml⁻¹. The number of nodules formed by R. meliloti and nodulation percentage were 7.5 per plant and 70% respectively at inoculum levels of 10³ cells ml⁻¹ and 0.1 nodules per plant and 10% nodulation at inoculums level of 10² cells ml⁻¹. A cross streak test on agar media showed that the growth of C. insidiosum and F. oxysporum or V. alboatrum were affected by the presence of one another. Similar observations were made with R. meliloti and the fungal pathogens. Glasshouse experiments showed that incidence of bacterial wilt infection was lessened when R. meliloti were present in the inoculation mixture with C. insidiosum. At inoculums level of 10⁷ cells ml⁻¹ of C. insidiosum, infection decreased when the concentration of rhizobia in the mixture increased. C. insidiosum, F. oxysporum, and V. alboatrum all caused reductions in nodule number and forage dry weight when inoculated alone. Mixed infections of C. insidiosum with either fungal pathogen caused even further reductions. Toxin isolated from liquid culture of C. insidiosum inhibited the growth of R. meliloti on an agar medium and reduced the number of nodules formed on lucerne plants. The FA technique showed that growth of C. insidiosum was less and the pathogen progressed slower in the xylem vessel elements of bacterial wilt resistant variety Saranac than susceptible variety Wairau.

Two field trials were conducted at Crop Research Division, D.S.I.R., Lincoln, in the 1971/72 and 1972/73 seasons to examine ways of reducing the cost of establishing lucerne and reducing the loss of production in the first year. Trial I examined the effect of a spring sown barley (cv Zephyr) cover crop at three seeding rates, 0, 56 and 112 kg/ha, on lucerne sown at the same time at three seeding rates, 2.9, 3.8 and 5.8 kg/ha with and without irrigation. The 1971/72 season was drier than normal but lucerne establishment was unaltered by a cover crop regardless of seeding rate or irrigation treatment. The overall establishment was high with 73% of viable seeds being present as plants after the first winter at a seeding rate of 2.9 kg/ha. Lucerne production at the harvest cut was reduced by barley sown at 56 kg/ha to 21% and 49% non-irrigated and irrigated respectively and remained depressed for two further cuts non-irrigated but yield was restored by the following cut irrigated. Irrigation increased the first season lucerne yield by 100% clear seeded and 194% with 56 kg/ha of barley and lucerne seeding rate had no effect on lucerne yield. Undersown lucerne at all seeding rates reduced barley grain yield by 20% at the low barley seeding rate but lucerne at the high seeding rate only reduced grain yield at the high barley seeding rate. Trial II examined the effect of a winter sown wheat (cv Arawa) and a spring sown barley (cv Research) on the establishment and early yield of lucerne sown in the spring at three seeding rates (2.25, 9.0, 35.0 kg/ha). The influence of cover crop competition was further examined by sowing the cover crops at three seeding rates (56, 112 and 224 kg/ha), and by simulated grazing of the wheat and the use of a herbicide on the barley and clear seeded lucerne. The 1972/73 season was drier than normal and although percentage lucerne establishment decreased as the lucerne seeding rate increased, no cover crop treatments effected lucerne establishment. Overall establishment was low with 36.5% of viable seed being present as plants after harvest at the 2.25 kg/ha seeding rate. Lucerne production at the harvest cut was low and was reduced markedly by cover crop treatments and height, root weight and root weight distribution data suggested that wheat had a greater effect than barley. No further cuts w.ere taken in the establishment season but cover crops continued to reduce lucerne yields in four cuts, in the following season. The use of spring sown cereal cover crops for lucerne establishment can be justified on the basis of providing a more profitable return, provided the establishment year production of clear seeded lucerne is low but cannot be justified on the basis of weed suppression. Lucerne seeding rates of 2.25 and 2.9 kg/ha produced stands with at least 30 plants/m² which was a density sufficient for maximum production.

A dryland lucerne (Medicago sativa L.) grazing trial to compare cv Wairau and a winter-active cultivar, Rere, was conducted over a two and a half year period using mixed aged ewes. Four cool-season grazing treatments were imposed; no cool-season grazing, grazing in mid-June, a short spelling in spring and a combination of the mid-June grazing and short spelling in spring. During the warm season, the trial was given one management of 10 days grazing with 42 days spelling. In the first season, there was no difference in total dry matter yield of Wairau and Rere. The winter-active, Rere, out yielded Wairau during part of the cool-season, while Wairau outproduced Rere in early spring. By the third season, however, the annual dry matter yield of Wairau and Rere was 10,820 and 7,300 kg DM ha⁻¹ respectively. This difference in yield was due, in part, to the appearance of crown rot in the final season. Over 60% of Rere plants were affected by crown rot, compared to only 3% of Wairau plants. The incidence of crown rot in Rere was increased by a further 20% when grazed in mid-June, although this grazing treatment did not affect the susceptibility of Wairau to crown rot. Rere was also 90% higher in weed content than Wairau when both had been grazed in June. However, grazing either cultivar in June removed overwintering aphids and resulted in lower spring aphid numbers than in lucerne spelled over June. This difference was only significant until aphid flights in November. The short spelling in spring depressed subsequent yields of both cultivars until mid-January, although aphid numbers on lucerne subjected to this short spelling treatment remained below 2 per stem until aphid flights in November. In contrast, lucerne given a longer spelling in spring had up to 90 aphids per stem prior to aphid flights. This spring grazing treatment did not affect the incidence of crown rot in either cultivar. After two seasons, combination of the June grazing with the short spelling in spring reduced the dry matter yield of both Wairau and Rere by 40% compared to lucerne with no cool-season treatment. Two other trials were established, with the lucerne cultivars Saranac, Wairau, Rere and Matador, which range from low to high winter-activity in that order. There was strong evidence from these trials that the depressive effect of June grazing on the yield of Rere in spring was due to a reduction in root reserves required for spring regrowth. Spring root reserves of the winter-active cultivars (Rere, Matador) were lower after the June defoliation than less winter-active cultivars (Saranac, Wairau) because of the greater use of reserves made by winter-actives in the cool-season utilisation of cool-season activity and high spring production thus proved to be incompatible.

Two experiments were carried out to investigate the effects of some management practices on the establishment of perennial grasses (‘Grasslands Matua’ prairie grass, ‘Grasslands Maru’ phalaris and ‘Grasslands Nui’ perennial ryegrass) overdrilled into a five year old lucerne sward and the subsequent impact of the introduced grasses on sward composition and productivity. The first experiment (24 Nov. ’89 to 26 Feb. ’90) evaluated the effects of contrasting levels of lucerne cover (open vs. closed) on initial establishment and seedling growth of perennial grasses overdrilled into lucerne. Lucerne cover had no effect on the start, rate and extent of germination. However, 90 days after sowing (DAS) the closed lucerne canopy reduced seedling survival from 81 to 28% in Nui, from 81 to 37% in Maru, but had no effect on Matua survival. The closed lucerne canopy significantly reduced all attributes of grass seedling growth examined: dry matter (DM) per plant was reduced to 4% (from 1620 to 70 mg), leaf area (LA) per plant was reduced from 98 to 6 cm². Root length density was reduced from 4.8 to 2.2 km m⁻³. The number of tillers per plant of Matua, Maru and under the closed lucerne canopy was reduced to 30, 20 and 13% of the open canopy, respectively. The second experiment (11 Sep.’89 to 24 Oct.’90) examined the effects of some management practices (rainfed vs. irrigation; and untreated vs. 4 g a.i. thiram kg⁻¹ seed) on the establishment of direct drilled grasses and subsequent impact of introduced species on sward composition and productivity. Irrigation increased, though not significantly, seedling emergence from 357 to 424 m⁻². Fungicide had no effect on seedling emergence. Seedling survival 247 DAS was similar under both moisture regimes (~ 11% of maximum emergence). Seedling establishment at 247 DAS was 9% for both Maru and Nui and 17% for Matua. Fungicide seed treatment had no effect on seedling survival. Seedling growth assessment from the open and closed canopies at 40 DAS indicated that canopy closure had more influence on seedling growth than irrigation regime. Under the closed canopy DM per plant was similar whether irrigated or not. Under the open canopy, however, growth increased four to five fold without and with irrigation, respectively. Irrigation doubled leaf area per plant of Maru and Nui but had no effect on Matua. On the other hand there was a 3-, 4-, and 5-fold response in LA plane⁻¹ to canopy closure in Nui, Matua, and Maru, respectively. Four months after sowing, irrigation increased root growth, in the top 20 cm of soil, from 314 to 420 g DM m⁻³. Matua, Maru, and Nui had 442,358, and 303 g DMm⁻³. There was an indication that the performance of the grasses under irrigation during late summer to early autumn was related to the size of their root system. During the establishment year (Sep. ’89 to June ’90) cumulative yield under rainfed conditions was 170,220 and 970 kg DM ha⁻¹ for Nui, Maru, and Matua, respectively. Irrigation increased these yields to 800, 2700, and 4960 kg DM ha⁻¹, respectively. Under dryland conditions, all grasses had no significant effect on lucerne yield. Under irrigation, however, Matua reduced lucerne yield from 15350 to 11210 kg DM ha⁻¹; Maru reduced lucerne yield from 15350 to 13100 kg DM ha⁻¹. Nui had poor persistence and performance and did not affect lucerne yield. During the establishment year total sward yield averaged 12.3 t DM ha⁻¹ and was not affected by the grasses. Under irrigation during late summer to early autumn yield differences were related to the amount of irradiance intercepted by each component of the mixture. During the 1990 winter Matua under both moisture regimes and Maru from the irrigated plots significantly reduced weed and bare ground cover whereas Nui from the rainfed and irrigated plots and Maru from the rainfed plots had no impact on both weed and bare ground cover. The lucerne component of the mixtures was, however, similar in all species-irrigation treatment combinations. In the first spring following the establishment year, total DM yield from the rainfed plots was 16% higher than that from irrigated plots and reduced weed yield to 17% of the pure lucerne plot but it reduced the lucerne yield to 55% of the pure lucerne plot yield. The total yield of the Matua/lucerne mixture was, however, 23% more than the pure lucerne yield. Maru while reducing weed content to 42% of pure lucerne plots, did not affect the yield of the lucerne component. The results suggested that Matua appeared to competitively exclude the lucerne which initially was a vigorous stand. It is thus suggested that overdrilling lucerne, while it maintains complete ground cover may not be beneficial. Maru, however, may be a suitable associate if the lucerne exhibits gapping at the normal time of harvest.

Two studies were conducted on pure WL320 lucerne, lucerne/ 'Grasslands Matua' prairie grass and lucerne/'Grasslands Maru' phalaris mixtures at Lincoln University, between April 1988 and September 1989. The 0.042 ha plots were sown in early November 1986 into a Templeton fine sandy loam soil. The weather during the experimental period was very dry during spring and summer, but was warm and moist in winter. The first study investigated the effects of short (3-7 days) and long (6-14 days) sheep grazing durations on the three pastures. Short grazing resulted in 15% more total yield from six grazing cycles (15000 versus 13000 kg DMha⁻¹). There were no significant differences in total yields between the pastures. Total lucerne yields were 20% higher under short duration for pure lucerne and lucerne/phalaris, but were not different for lucerne/prairie grass. The total yield of prairie grass under short grazing (9160 kg DMha⁻¹) was 36% higher than under long grazing (6740 kg DMha⁻¹), but phalaris yields were similar at 2570 and 2320 kg DMha⁻¹ respectively for short and long grazing. The results indicated that short grazing duration was superior to long grazing duration, when rotation lengths were equal. The second experiment compared the performance of lucerne with the lucerne/grass mixtures in shallow (80 cm) soils on mini-plots of 8 m² selected from areas within the grazed plots. Water use from the soil was measured using a neutron probe. Root cores were sampled in May 1989 from deep soil plots. On the shallow soils total yields from pure lucerne and lucerne/phalaris were similar (11700 and 11400 kg DMha⁻¹) and were significantly higher than the lucerne/prairie yield of 8500 kg DMha⁻¹. Pure lucerne and lucerne/prairie yields were 17% and 45% higher in the deep soils than in the shallow soils, but in both soil types, lucerne/phalaris yields were similar. Winter yields from pure lucerne and lucerne/prairie were similar on both soils (3000 kg DMha⁻¹), but lucerne/phalaris produced 27% more on shallow (3700 kg DMha⁻¹) than on deep soils (2900 kg DMha⁻¹). Total lucerne production in lucerne/phalaris of 9900 kg DMha⁻¹ was not affected by soil depth but in lucerne/prairie, lucerne yield was 57% less on shallow (3950 kg DMha⁻¹) than on deep soils (9100 kg DMha⁻¹). Total yields from prairie grass and phalaris were higher by 30% and 10% respectively on shallow soils. Total water use was not different between the pastures at 384, 376 and 375 mm respectively for lucerne, lucerne/prairie and lucerne/phalaris, but was 10% higher on deep soils (398 mm) than on shallow soils (360 mm). Seasonal water use was similar for all three pastures and was higher in deep soils by 39% in spring and 20% in early summer. All three pastures showed similar water use efficiency (WUE) in deep soils but in shallow soils, the WUE of lucerne/prairie was significantly lower (10-13 kg DMmm⁻¹ H₂O) than either the lucerne or the lucerne/phalaris mixture (18-24 kg DMmm⁻¹H₂O). Pure lucerne extracted water from greater depth than the mixtures. Under all pastures, water was extracted until soil volumetric water content reached 10%. Pure prairie grass and phalaris had 70% and 50% of their roots in the top 20 cm, and roots of both grasses were traced to 70 cm depth. In the top 20 cm lucerne root yield of lucerne/prairie was only 60% (4 kg DMm⁻³) compared to 88% (9 kg DMm⁻³) of lucerne/phalaris at that depth. The results showed that under the extremely dry conditions of 1988/89, there was no advantage in pasture yield of lucerne/grass mixtures over pure lucerne. However the inclusion of phalaris in shallow soils and prairie grass in deep soils provided effective weed control and good cool season growth without adversely affecting pasture productivity.

A grazing study was conducted on irrigated Wairau lucerne (Medicago sativa L.) during summer and autumn using mixed aged ewes. Two treatments were imposed; grazing durations of 3 days and 12 days with similar spells to the early flower stage of growth between successive grazings. This was carried out over two grazing and regrowth cycles. Weekly measurements of regrowth following grazing showed that for both periods, regrowth following a 12 day grazing duration was more rapid than regrowth following a 3 day duration, in the early stages. The superiority developed initially was maintained until the early flower stage. Yields of dry matter were 4170 kg DM/ha and 2435 kg DM/ha for each regrowth period following grazing for 12 days. Corresponding yields following grazing for 3 days were 4000 kg DM/ha and 2350 kg DM/ha. Concurrent with the second grazing and regrowth periods a cutting trial was set up to simulate the long and short grazing durations. The long simulated grazing duration depressed dry matter yield at the early flower stage compared with the short simulated duration, whereas actual grazing demonstrated the superiority in yield of a long duration. Therefore, conclusions that would have been drawn from a cutting trial alone were different from those drawn from the grazing trial, demonstrating that care must be exercised in interpreting results of simulated grazing trials. Dry matter yields of regrowth following cutting were considerably lower than the yields obtained following grazing. The patterns of grazing in a long and a short grazing duration were compared. Sheep grazing for 12 days ate more of the available dry matter than sheep grazing for 3 days. The difference occurred in the early stage of a long grating duration when the lower stocking rate resulted in less competition for the most palatable plant parts, the apices and leaves. For both grazing durations the rate of dry matter consumption declined as the duration progressed because of a decline in digestibility and increase in the percentage of unpalatable stem material in the available dry matter. "In-vitro" digestibility measurements showed that the digestibility of the leaf less stalk at the end of grazing for 12 days was equivalent to lucerne hay, having declined 20% since the start of grazing. The digestibility of lucerne declined only 10% during regrowth to the late flower stage in autumn. Observations during regrowth following grazing indicated a stimulatory effect on plant growth and vigour of dung and/or urine. This was not apparent on areas receiving cutting management. Measurements of plant nitrogen in such areas showed that levels had been raised by animal returns. This is discussed together with other work indicating that nitrogen fixation in lucerne may not always be adequate for maximum growth.

In order to improve the level of understanding of the reproductive seasonality and flight potential of Silona discoideus Gyllenhal, samples were taken at regular intervals from lucerne (Medicago saliva L.) stands throughout the 1982-83 and 1983-84 seasons. Changes in the weevil's reproductive condition and flight muscle morphology were monitored by dissection. This work confirmed that the univoltine adult weevils entered aestivatory diapause some three to four weeks after the eclosion of the new generation in December-January each year. Flight to aestivation sites remote from the host crop, was preceded by a vigorous feeding phase during which time the indirect flight muscles developed. The existence of aestivatory diapause was further supported by an observed suppression of reproductive development, absence of mating, increase in fat content of the weevils and changes in the proportional representation of several fatty acid constituents of the total fatty acid complement. During aestivation there were also changes in the indirect flight musculature of S. discoideus. Following flight to aestivation sites, the indirect flight muscles atrophied. Subsequently, the muscles regenerated, enabling reinfestation flight to the lucerne to commence. Aestivatory diapause was found to last some seven to eight weeks. However, in late eclosing weevils, the predetermined time period for diapause development was not always completed before cool winter conditions limited flight muscle regeneration. Hence, throughout this study, some S. discoideus were able to be located at aestivation sites. Laboratory experiments revealed that depleted metabolic reserves in these weevils ultimately made muscle regeneration impossible irrespective of temperature, thus precluding a spring invasion of lucerne by these laggard aestivators. Generally reproductive activity in adult S. discoideus was found to commence with the successful termination of aestivatory diapause. Oviductal eggs were located in the lateral oviducts from mid March onwards, while egg populations became apparent in the field in mid April. The importance of the winter months for larval establishment of S. discoideus was determined in Canterbury. Chlorpyrifos applied at 0.6kg a.i. ha⁻¹ on lucerne field plots was found to effectively eliminate adult S. discoideus populations but left their eggs unaffected. A field experiment was thus established whereby populations of reproductive weevils were eliminated at 5-weekly intervals throughout the 1983 egg laying period. The importance of the different egg laying time spans was thus determined by sampling the resultant larval populations in spring. There were found to be two main peliods of S. discoideus larval recruitment: the first in early autumn and the second, more important in spring. The current recommendation to prevent S. discoideus larval damage to lucerne is to spray insecticide against the weevils in late autumn, subsequent to the post-aestivatory reinfestation flights but before the onset of the bulk of their reproductive efforts. The present findings suggest an insecticide spray applied to susceptible stands as late as mid August could usefully reduce spring larval populations in most years. Analysis of head capsule widths of both field-collected and laboratory-hatched s. discoideus larvae confirmed that this species possesses five larval instars. Probit analysis allowed for more accurate statistical definition of the head capsule size range of each particular instar, despite some size overlap between successive instars. Such accurate definition of instar sizes subsequently pelmitted detection of both intra- and interseasonal variations in the within-instar head capsule sizes of S. discoideus larvae duling each of the 1982-83, 1983-84 and 1984-85 seasons. Such variations were attdbuted to intra- and interseasonal variations in the rhizobial roqt nodule availability/accessibility to the larvae, i.e., changes in food availability.

For many years plant breeders have been trying to develop a persistent pasture legume for the Great Plains of the U.S.A. and Canada, and more recently for the drier areas of Australia and New Zealand. Of the legumes tested in the Prairie region, lucerne has proved to be the most productive and persistent. However, even the hardiest of the common varieties do not persist through a series of dry years. Recently, creeping-rooted lucerne was introduced into New Zealand as a possible variety for use in the drier areas. In its introduced form it was found to be too low producing to become an important economic variety but by crossing with the higher producing Medicago Sativa types and selecting for creeping habit it is hoped to evolve better strains. It is felt that higher producing strains of creeping-rooted lucerne could have an important place in the drier areas of Canterbury and Otago, and possibly solve the legume problem in the high country, if it could be successfully established. Creeping-rooted lucerne has horizontal roots from 4-6" underground which are capable of producing adventitious shoots at intervals along their length. These serial shoots either develop a root system of their own or rely on the parent root. (Fig. 2, Page 3) The plants spread readily without moist surface conditions. The phenomenon of adventitious shoot formation on roots is an interesting one. As Torrey (1958) points out, “the capacity for endogenous bud formation is a distinctive character of relatively few groups of plants, and involves a mechanism about which we have very little physiological knowledge”. Previous workers have dealt almost entirely with shoot production on root segments isolated from the parent plant. Although this is a convenient method for testing the potential of root segments to produce adventitious stems, and studying their formation, its value is somewhat limited. It was felt that a wider understanding of the problem would be obtained if adventitious shoot production on the roots was studied in relation to the physiology of the whole plant. This study therefore deals with the effects of environment on the growth of the plant as a whole, with particular reference to the creeping-rooted habit.

The results of testing lucerne for resistance to the stem nematode Ditylenchus dipsaci are recorded. A major part of the study was given to perfecting a technique for growing lucerne seedlings for inoculation with nematodes. Two techniques were tried and filter paper rolls were found to be superior to soil flats. The technique of growing seedlings in filter paper rolls and inoculating them with nematodes was used to produce a rating of the resistance for the New Zealand variety of lucerne, Wairau, when compared with a known resistant variety, Washoe and a susceptible variety, Grimm. From the results obtained Wairau must be classified as a susceptible variety. It appears that a selection within a variety of lucerne for resistance to Ditylenchus dipsaci could be made by eliminating the swollen plants. However, to obtain the relative resistances of several varieties, it would be necessary to obtain a measure of the reproductive rates of the nematodes in the plants.

Over recent years, there has been a considerable increase in the area of lucerne grown in New Zealand, particularly that used for grazing. The total for conservation and grazing increased from 19,000 ha in 1947 to about 145,000 ha in 1968 (White, 1968) and it is expected that the figure will reach 200,000 ha by 1973. In the same manner, the proportion used for grazing rose from less than 10% in 1947 to more than 40% in 1966 and it is anticipated that this trend will continue into the foreseeable future. More recently, the importance of pure lucerne, especially in early spring, has become greater following the general trend of intensifying animal production through early weaning of lambs (Jagusch et al., 1970; Jagusch et al., 1971; Joyce et al., 1972; McConnell, 1972). One of the main aims of the trials reported in this thesis is to verify Professor K.F. O'Connor's theory that removing the apices of immature lucerne plants stimulates the growth of crown buds and at the same time enhances the growth of axillary shoots on the upper portions of the stubble. Results of such findings could prove to be valuable in future in view of the reported increase in the number of young lambs weaned on to lucerne pastures and noting also that the grazing behaviour of these animals differs markedly from that of the older stock. Further, it has been shown that leafy tops (stem apices) and immature lucerne or cool season growth are high in feed quality (Bailey et al., 1970; Jagusch et al., 1970; Smith, 1970; Jagusch et al., 1971). Although a few workers (O'Connor, 1970; Peart, 1970; McKinney et al., 1972) have attempted to provide balanced data on the performance of both animals and pastures in a grazing system, little is known about the individual plant responses to varying grazing behaviour, particularly its reactions to different periods of grazing and spelling. Lack of information on the above aspects prompted the cutting treatments described in this thesis to be designed and the experiments carried out. It is hoped that the principle of optimal growth and production or other relevant information discovered from the cutting trials could gain useful applications to a real grazing situation.

The importance of pasture and pasture plants in soil conservation in world agriculture cannot be gainsaid. Unfortunately pasture lands heir to a large number of problems such as ravages of serious erosion in most parts of the world. Thus grassland management involves a constant struggle against these forces of destruction such as soil slip, slump, gully erosion, debris avalanche, and so on. Although there has been an immense amount of work done on this subject, resulting in the accumulation of a large and varied literature on the subject, little effort seems to have been made especially in New Zealand to bring this widely scattered material together so as to present a more or less complete picture of the role of grass in soil conservation before those concerned with its cultivation, management, and improvement. This has resulted in an under-estimation of the actual alarming state of affairs as people are either apt to overlook the danger till it really assumes a serious form, thinking that if the danger is absent today it will remain absent even tomorrow. In the present work an attempt has been made at the synthesis of our current knowledge about pasture and pasture research in New Zealand with special reference to the role of pasture and pasture plants in soil conservation. Briefly its applicability has been discussed in such countries as U.S.A., Canada, South Africa, Australia, and its possible applications to the problem of soil conservation in India. The necessity of this work was felt because of the fact that there is no comprehensive book dealing with this subject exclusively to cater for the needs of both the research worker as the scientist and the farmer as the practitioner, With the result that if anyone wants to have some definite information on the topic, he has frequently to perform the Herculean task of exploring a veritable ocean of literature on the subject. In the first place, everyone is not so favourably situated as to have an unrestricted access to all the relevant literature required; secondly, even assuming that he has such access, the time and energy involved in such a search is almost breath-taking; and lastly, very often he is confronted with a maze of conflicting views which make his task really stupendous. The object of this work is to obviate such difficulties as far as practicable. Throughout the preparation of these pages, the aim has been to condense the extensive literature available on the subject to the size of a handy volume for the benefit of all who are interested in the topic. The writer would have liked to be a little more critical than he has generally been throughout the survey, but this could not be achieved owing to the limitations of space and time. The magnitude of the task before the writer can be guessed from the fact that the whole work has been done only with one year's background of New Zealand farming. The bibliography will be found to include a few more references than would be warranted by the contents. This has been done to provide readers with access to relevant literature if they wish to have more information on any particular aspect. Attempts have been made to quote authorities for all information given but this has not been possible in every case. Where quotations have been used they have often been summarised or paraphrased and in such cases inverted commas have not been used. The appendix contains an account of small research project conducted by the writer himself. This work on the effect of varying day lengths and temperature on six different strains of lucerne was undertaken to gain an insight into original work in plant science; and though the field covered is extremely narrow, it has nevertheless served the objective admirably.

The work reported here is an evaluation of leaching losses from Medicago sativa. Lucerne plants subjected to rain, dew or mist lose nutrients by leaching, the extent of nutrient loss being influenced by a number of factors, some of which are considered in the text. In this work phosphorus and sulphur were chosen for study. Radioisotopes of these elements can be safely handled and easily and accurately detected in the laboratory; both elements are of major importance in plant nutrition; many New Zealand soils have limiting supplies of these elements and a knowledge of the part that leaching may play in plant-soil relationships would therefore be important. Section 1 considers the physiological factors which may influence the magnitude of leaching. Section 2 considers leaching from lucerne hay at varying stages of moisture loss and under the influence of smooth-roller hay conditioning.

Insects and other pests of Lucerne in New Zealand have received only spasmodic and cursory attention in the past (see Section 2.1). As a result, basic and applied information on insects associated with lucerne in New Zealand is limited. It is important to make an overall assessment of the fauna associated with Lucerne to overcome the previous deficiency of information, particularly to establish which species are pests (see Chapters 4. 9, 10 and 12). The need for this information becomes more important as the increase in lucerne acreage continues, and realization of the value of lucerne grows (see Chapter 3). A quantitative survey was considered the best way of making such an overall assessment of the fauna, because it can reveal the abundance and distribution of single species with known precision (see Chapter 6). Pilot sampling for survey purposes can also assist in the formulation of optimal sampling plans for particular species or groups. The primary aims of this study were as follows: 1. To find out what species of arthropods and other animals are associated with Lucerne in the South Island (see Chapter 7). 2. To attempt a preliminary economic assessment of Lucerne posts in New Zealand (see Chapters 4, 5, and 7 to 12). 3. To indicate optimal sampling plans for the more abundant insects associated with Lucerne (see Chapters 6 and 8). 4. To examine some basic aspects of Lucerne seed production (see Chapter 11). Subsidiary aims to these primary aims are given in Chapters 7 to 11. Obviously with such multipurpose and comprehensive aims it was difficult to achieve conclusive results for all of them. It was felt that such a comprehensive study was best for a preliminary study, because of the absence of knowledge on insects associated with Lucerne and methods for measuring their populations. Consequently, the primary aims were to establish procedures to more specific studies in the future, and to provide on objective preliminary assessment of the fauna associated with Lucerne. In addition, it was felt that the value and scope of a single plant ecosystem study (see Chapter 2) could be demonstrated in conjunction with the primary aims. Hence a secondary aim of this study was to examine the problems of single crop ecosystem studies (see Chapters 2 and 12). The author’s philosophy is included in Chapters 2 and 6 because ecosystem studies and preliminary quantitative surveys are still in an early stage of development, so that clearcut methodology and theories have not been developed at the single plant level for associated faunal studies. The arrangement of Lucerne pests in Chapters 4 and 5, which is partly based on the results of this study, is primarily for agriculturalists rather than the common arrangement based on insect taxonomy, which was used by Pottinger and Macfarlane (1967) in a preliminary review of Lucerne pests in New Zealand. Otherwise, this thesis follows a conventional pattern of introduction (Part I), Review of Literature (Part II), Methods and Results (Part III), and conclusions (Part IV).

In recent years there has been an upsurge of interest in lucerne as a pasture plant for prazing, either sown alone of in a mixture. This has been accompanied by an increased interest in the effect of a higher proportion of M. falcate genes on the agronomic characteristics of lucerne. Since M. falcate is tolerant of more extreme environments than M. sativa, the infusion of M. falcate germ plasm into lucerne breeding programmes has conferred several advantages. The most notable of these has been the breeding of creeping lucernes, both rhizomatous and creeping rooted types having been developed. The available evidence suggests that not only are the creeping lucernes more tolerant of grazing than non-creeping cultivars, but they can be of special value in environments subjected to drought or extreme winter conditions. In an attempt to even out seasonal variability on light soils, several cultivars of lucerne were sown at Ashley Dene, in a mixed sward under grazing conditions in 1950 (Iversen 1965, 1967). Yields were increased from the 3500 lb D.M./acre normally achieved on this soil type, to 5000 lb D.M./acre, with reduced variability. Within five years the standard New Zealand cultivar (Wairau) had thinned to nothing, several other strains were little better, but Glutinosa, a rhizomatous creeper, persisted well and was still thriving after sixteen years (Iversen 1967). In 1954 the Plant Science Department at Lincoln College commenced an improvement programme with Glutinosa, aiming to produce a grazing type lucerne more suited to the Canterbury environment. The final selection of six parents was made in 1963 and the first seed obtained in 1964. Since then a number of field trials comparing College Glutinosa and Wairau in various environments, have shown that College Glutinosa can be a valuable alternative to Wairau, especially as a grazing plant (J.G.H. White, pers.comm). Little work has been done to investigate the responses of College Glutinosa to variations in environment or management systems. The present study was designed to obtain information on the normal pattern of crown development under simulated sward conditions. The pattern of shoot formation and development was also studied more fully under controlled environment conditions in growth cabinets.

In the course of these studies more questions have been posed than answered. It is clear that Lucerne production is being limited by low Cu levels in some areas. The relationships between yield and plant composition were rather different from those predicted from other studies; it is considered that this indicated that the present understanding of these relationships is far from adequate. A survey of Cu content of Lucerne growing on Eyre, Lismore and Wakanui soils showed the first two to supply just enough Cu for healthy plant growth at most sites. Wakanui soils grew Lucerne with more than adequate Cu contents. Variation within stands was not great, bulk sampling should give adequate information about the Cu status of any particular stand. Variation due to season was not marked. It is concluded that herbage Cu levels on all soils surveyed were sufficiently low to cause concern about animal health and that it is reasonable to assume similar or lower Cu levels in the other “old” shallow soils of Canterbury.

In the South Island of New Zealand there are more than ten million acres of tussock and danthonia grassland. Many research workers and farmers have shown that much of this area can be greatly improved by the introduction of clovers. However in the semi-arid (