Search

The abolishment of the European Union (EU) milk production quota in 2015 has led to rapid expansion of the dairy industry in Ireland. However, where access to additional land is limited, farmers must explore alternative or additional feeding strategies to improve milk production and enhance production efficiencies in an environmentally sustainable manner. With an increased emphasis on milk solids production (MS), the incorporation of white clover into the sward or supplementing dairy cows with total mixed ration (TMR) or partial mixed ration (PMR- pasture supplemented with TMR), may be viable options to improve animal performance over a grass-only diet. The aim of this thesis was to identify dietary strategies to optimise cow efficiency in temperate regions where grazed grass is generally the cheapest and most predominantly fed forage for milk production. Four feeding systems were compared as part of this study; cows grazing outdoors on perennial ryegrass (Lolium perenne L.; PRG) only swards receiving 250 kg nitrogen (N)/ha (GRASS), grass-white clover (Trifolium repens L.) swards receiving 150 kg N/ha (CLOVER), cows housed indoors and fed a total mixed ration in a 100% confinement system (TMR), and cows fed a combination of pasture and TMR (partial mixed ration; PMR). After calving each year, cows were assigned to one of the treatment groups based on calving date, lactation number, 2 week pre-experimental milk yield, and 2 week pre-experimental MS (kg of fat and protein) yield. Cows remained assigned to their treatment groups for the entire lactation in each year. The grazing treatments were stocked at 2.74 cows/ha. Both the GRASS and CLOVER treatment groups grew 14.1 t dry matter (DM)/ha despite an extra 100 kg N fertiliser being applied on the GRASS treatment. At an average white clover content of 21% over the duration of the study, the CLOVER cows produced an additional 18 kg of MS, which translated into an extra 51 kg MS/ha compared with the GRASS cows. The TMR system supported the greatest daily and cumulative milk and MS yield compared with the pasture-based diets. In terms of reproductive performance, the majority of parameters were not influenced by feed system with similar conception rate to first service, 21-day submission rate, conception rate to second service, and overall pregnancy rate. However, the CLOVER cows required six days less between calving and conception (93 days) compared with the GRASS and TMR cows who both required 99 days. Additionally, a greater proportion of CLOVER and TMR cows were pregnant to their first service. Bodyweight (BW) was significantly greater for the TMR fed cows compared with the pasture-based cows; the TMR cows were 29 kg heavier over the entire lactation (553.4 vs. 524.7 kg). Average daily milk and MS yield was greatest for the TMR cows (24.7 and 2.05 kg/cow, respectively), intermediate for the CLOVER cows (20.3 and 1.68 kg/cow, respectively), and lowest for the GRASS cows (19.1 and 1.56 kg/cow, respectively). This is a reflection of the total dry matter intake (TDMI; DMI) being significantly different between treatments. On average, the TMR consumed 2.30 kg/cow greater than the CLOVER cows, which consumed 1.03 kg/cow greater than the GRASS cows. Similarly, cows consuming the TMR diet, had a greater energy intake in comparison with the pasture-fed cows. Consequently, this influenced the production efficiencies. The TMR cows had greater amounts of energy available for milk production after maintenance and also had greater total energy available to produce 1 kg MS compared with the GRASS and CLOVER treatments. When the PMR system was incorporated into the study in 2020, the daily milk and MS production and cumulative totals (24.35, 2.09 and 6701, 584 kg/cow, respectively) were intermediate between the fully nutritionally balanced TMR system (25.60, 2.25 and 6988, 629 kg/cow, respectively) and the GRASS system (20.54, 1.77 and 5770, 526 kg/cow, respectively) over two full lactations, despite a similar DMI for the PMR and TMR (20.90 kg/cow). A similar trend to that for milk production efficiencies was also observed for DMI per 100 kg BW, fat and protein corrected milk (FPCM) per 100 kg per 100 kg BW, and MS per 100 kg BW. However, the PMR cows produced the lowest quantities of MS per kg total DMI which suggests that energy was being diverted from milk production towards walking and grazing behaviour when a combined pasture and TMR feeding system is implemented. While throughout this study, the TMR system supported the greatest milk and MS yield over the pasture-based diets, a consequence of this was significantly greater daily enteric methane (CH4) emissions and gross energy intake lost as CH4. The TMR cows emitted, on average over the five days of each of the three measurement periods (May, July, and September), 610.0 g CH4/cow, the CLOVER cows emitted 436.9 g CH4/cow, while the GRASS cows emitted 404.5 g CH4/cow. Additionally, the TMR feeding system supported the greatest gross energy intake losses as CH4, the CLOVER was intermediate and the GRASS was lowest (9.0 vs. 7.7% vs. 7.4%, respectively). The GRASS and CLOVER cows emitted similar quantities of CH4 in the May and July measurement periods, while in September the CLOVER cows emitted 67.56 g/cow per day more CH4 compared to the GRASS group. Nonetheless, both pasture-based treatments emitted a similar amount of CH4 when corrected for DMI (30.32 g/kg DMI). This research PhD study proves that milk production and its associated efficiencies can be improved by feeding a TMR or PMR, or by incorporating white clover into the sward compared that achieved by a pasture-only diet. Nonetheless, in temperate regions, grazed grass remains the cheapest feed source for milk production, any increases in animal performance obtained through alternative feeding strategies must be environmentally and economically viable.

The aim of this thesis was to investigate offsetting conventional nitrogen (N) inputs with more sustainable options in order to maintain or increase production. This was examined through four separate investigations, namely, 1) a retrospective analysis of the incorporation of the legume white clover (Trifolium repens L.; WC) into PRG swards within a high N input grazing dairy system (Chapter 4), 2) a systems experiment investigating PRG-only and PRG-WC swards at two levels of N fertiliser (150 and 250 kg N/ha; Chapter 5), 3) supplementing PRG diets with a rumen protected protein source compared to a rumen degradable protein source within a low CP concentrate supplement (Chapter 6) and 4) comparing herbage production of different N fertiliser types and rates (Chapter 7). Chapter 4 interrogated an existing database, comprised of results from a larger, overall systems experiment, that showed fluctuations in WC content in PRG swards within a high N input (250 kg N/ha, annually) grazing dairy system. In Chapter 5, a separate grazing experiment was established that compared both herbage and milk production of PRG-only and PRG-WC swards at 150 and 250 kg N/ha. Chapter 6 investigated the impact of supplementation of differing protein sources at low amounts on milk production and N use efficiency. Finally, Chapter 7 investigated N fertiliser type and rate on PRG herbage production in a rotational temperate grazing system through a comparison of calcium ammonium nitrate (CAN), urea + N-(n-butyl) thiophosphoric triamide (urea + NBPT) and urea at 150 and 250 kg N/ha per year. In summary, this thesis explored varying options within intensively grazed temperate dairy systems to offset and protect the N inputs while maintaining or perhaps increasing milk production.

Drought is one of the abiotic stresses with more effect on crop productivity world-wide. Understanding plants' response to drought is fundamental to produce drought tolerant varieties. Marker-assisted selection (MAS) helps reduce the time needed to breed new varieties by using markers associated with the trait being introduced. The work presented in this thesis is a first step in the identification of the molecular markers associated to drought response and recovery in white clover (Trifolium repens). The LemnaTec high-throughput phenotyping system was used to develop a total of 12 indexes that allowed us to phenotype the response to drought of the Stolon Mapping Population (SoMP). An integrated linkage map of the parents of the SoMP family was constructed from SNPs. The integrated map generated has 312 markers over 3211.283 cM on 14 linkage groups (LGs). The analysis of QTL was performed using both linkage analysis and association analysis. By linkage analysis several clustered QTLs were found on LG 8b suggesting this LG is a hotspot for drought response/recovery on white clover, especially in the control of greenness/yellowness. Both the indexes developed, and the integrated linkage map created during this thesis will help in the advancement of the knowledge and understanding of the molecular mechanisms of the drought response in white clover, helping in future QTL linkage analysis and MAS of drought tolerant white clover varieties.

Incorporating white clover into intensive dairy production systems has the potential to increase their sustainability through increased milk output and reduced nitrogen (N) input. To support intensive stocking rates (<2.75 cows/ha) some inorganic N will still be required to support herbage production, particularly in the spring. However, sward white clover content, and hence N fixation has been shown to be suppressed by applications of inorganic N. In this thesis, the effects of varying N application rates on grass-clover swards as well as the effect of those swards on overall herbage production, dairy cow production, dairy cow rumen function and overall farm gate N balance were investigated. White clover inclusion increased herbage production at fertiliser application rates of up to and including 180 kg N/ha/year in the plotbased study described in chapter 4 of this thesis. The results from this demonstrated the negative relationship between increasing N fertiliser application rates and sward white clover content with sward white clover content decreasing as N application rate increased from 0 kg N/ha/year to 240 kg N/ha/year. Clover inclusion was found to increase milk solids production per cow and per hectare. There was no difference in cumulative herbage production between grass-clover swards receiving 150 kg N/ha/year and grass-only and grass-white clover swards receiving 250 kg N/ha/year. This indicated that the N fixing ability of the grass-white clover swards was supporting sward productivity at the lower N regime. Clover inclusion in the diet significantly altered the rumen function of the dairy cows with increased rumen ammonia and total VFA concentrations in the rumen of the grass-clover fed cows over those of the grass only and total mixed ration (TMR) fed cows. The farm gate N balance outlined the environmental benefits of white clover inclusion in grazed swards resulting in improved overall N use efficiency (NUE) and reduced N surpluses in the grass-white clover swards compared to grass-only swards. Overall, this thesis provides evidence that white clover can be successfully incorporated into intensive dairy production systems with benefits in terms of farm output and improved on-farm sustainability, though with both positive and potentially some negative environmental implications. The overall outcome of the project was to provide evidence that it is possible to maintain and benefit from clover inclusion in grass swards that are intensively managed at high N fertilizer rates, if grazing management is of a high standard with good control of pre- and post-grazing heights. The project also identified some knowledge gaps, particularly around achieving good spring sward productivity, differing environmental implications and longer-term stability of the clover system, that provide clear objectives for future research in this area.

New Zealand dairy farmers rely primarily on swards based on perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.) for feeding dairy cows. The direct-grazed, low-cost production system that these pastures enable provides the New Zealand dairy industry with a competitive advantage for products in the international market. As a result, the industry contributes significantly to New Zealand’s export earnings and national gross domestic product. However, farmers in the upper North Island have expressed concern about the poor productivity of their perennial ryegrass pastures beyond three years post-sowing and have linked this to more-frequent occurrence of drier and warmer conditions for pasture growth during summer and autumn. Recent research has shown that even pastures sown with new ryegrass cultivars inoculated with the most-advanced endophyte strains cannot sustain high productivity under the severe climatic conditions and increased insect pest burdens being experienced in many parts of the Waikato, Bay of Plenty and Northland. In some cases, this is due to the death of ryegrass plants and ingress of lower-producing weed species, while in other cases reasonable ryegrass plant density can be sustained, but growth rates are poor. There is limited information on the plant population and phenological factors associated with these changes, making it difficult for farmers, agronomists and plant breeders to develop effective solutions. The study described in this thesis investigated the differences in phenotypic trait expression of perennial ryegrass plants removed from 10-year old pastures grazed by dairy cows in the Waikato (‘survivor’ plants) versus plants grown from the original seed lines that were used to sow the pastures and had, therefore, not been subject to any environmental selection pressure (‘reference’ plants). The study aimed to help inform future plant breeding efforts to utilise phenotypic changes and reselect for traits that are associated with better plant survival. Three populations were compared: reference plants; survivors from pastures rotationally grazed by dairy cows since they were established in autumn 2011 (‘control’); and survivors from pastures managed as per the control treatment, but where grazing was deferred for ~ 120 days from mid spring to mid-summer to promote new seedling recruitment two years before plants were collected from the field (‘deferred’). Four perennial ryegrass cultivars were included in the design: Nui standard endophyte (SE); Commando AR37, a mid-season heading diploid; Alto AR37 a late season heading diploid; and Halo AR37 a very late heading tetraploid. Thus, 12 populations were compared. Initially, there were 60 plants per population giving a total of 720 plants which were grown individually in pots at Lincoln University. The first phase was a pilot study during the early acclimatisation period where progress towards equalisation of plant growth across the populations was tracked. Biomass accumulation and reproductive development were assessed from late February to late April. Substantial flowering activity was recorded in the control and deferred populations for all cultivars, but not in the reference population, illustrating the expected carry-over effects related to differences in plant source (clonal fragments from the field versus seed-derived for the reference population). Biomass accumulation was initially lower in the reference population but there were no differences between populations by day 105 when plants had tillered-out to a similar degree. All plants were then analysed for the presence or absence of endophyte and, where endophyte was present, its genetic identity (strain). Plants with no endophyte, or with an endophyte strain that was not AR37 (or not SE, for Nui), were discarded and plant numbers rebalanced to 30 plants per population that were confirmed as ‘true-to-type’ for the original host genotype – endophyte strain combination sown in 2011. Four further populations of control plants (one for each cultivar; total 240 plants) were available for endophyte analysis from a site in Canterbury where the same experimental design and seed lines as used in the Waikato was established at the same time. The grazing deferral treatment was not implemented at this site. For Waikato, the percentage of plants that were true-to-type was significantly higher in the reference population (82%) than the control and deferred populations (69-71%). Overall, Commando had significantly lower percentage true-to-type than the other cultivars (60% versus 78-82%), with a high presence of standard endophyte (22%) indicating contamination by volunteer ryegrass plants and poorer persistence of the original population. For the Canterbury control populations, mean true-to-type percentages were 85-98% across the four cultivars, i.e. generally higher than for the original seed lines. The second phase was a detailed phenological study, where plant, leaf and tiller characteristics were measured on the true-to-type plants only, and flowering observations were conducted for 70 days starting in mid-October. Phenotypic measurements included: leaf weight, length, width, area, specific leaf area and thickness; pseudostem weight, length and width; and visual scores of rust, tillering and growth habit. Control plants had significantly lower leaf dry weight, shorter leaves and smaller lamina area than the reference plants, with no interaction between population source and cultivar. Deferred populations did not differ from reference populations in these traits, indicating grazing deferral ‘re- set’ the plant phenotype back from the control state to its original state, at least for the first 18 months after deferral. There were no differences between populations in pseudostem characteristics, growth habit or tiller density scores, but reference populations showed a trend towards higher incidence of rust compared with control and deferred. A significant shift toward later flowering in the control population compared with the reference population was found for Nui, Commando and Alto, but not for Halo. In summary, differentiation toward smaller leaves, greater disease resistance, and later flowering was observed after nearly ten years under grazing in the Waikato populations when potential contaminants were excluded and any effects due to endophyte presence/absence or variant strains were removed. It appears that the shift towards longer leaves can be reversed by deferring grazing to allow recruitment of new seedlings into the population. In contrast, shifts in rust resistance and flowering in control populations were not reversed as a result of spring/summer grazing deferral. Further investigation of the genetic factors controlling the reversibility of leaf traits could help identify breeding objectives for maintaining the persistence of yield advantages in ryegrass pastures. Future studies should also examine possible long-term benefits of this management practice to bring about phenotypic change for yield improvements and consider the duration of possible carry-over effects.

This investigation compared the seed yield of two cultivars (‘Legacy’, ‘Quartz’) of white clover and related these to morphological development and light interceptance in different row spacings (30, 45, 60 cm) and light regimes (north and south). This experiment at AgResearch, Lincoln in Canterbury, New Zealand ran from 17th March 2021 until a final (mechanical) harvest on the 25th February 2022. The development of tagged stolons was monitored weekly. Light interception and ground cover were recorded weekly until canopy closure. In addition, three destructive harvests were taken to capture key morphological characteristics of stolon length and seed yield from stolon, hand (quadrat) and mechanical harvests. Above average rainfall immediately prior to harvest resulted in low yields, which was consistent with regional commercial white clover seed yields. ‘Legacy’ produced more seed (792 kg/ha) than ‘Quartz’ (552 kg/ha) when harvested by hand. However, ‘Quartz’ may have had a higher seed potential because it had more juvenile buds and flowers present at desiccation compared with ‘Legacy’. Later desiccation of ‘Quartz’ would enable further flower ripeness and seed yield and it is hypothesised it may then have had seed yields comparable or higher than ‘Legacy’. The ‘Quartz’ phyllochron was 17.5°Cd less than ‘Legacy’ and it had a faster node appearance and stolon extension rate with more branches produced. This resulted in more leaves and total stolon length at stolon harvest compared with ‘Legacy’. An increase in row spacing or the sunny side of rows (north) may have resulted in microclimates, with increased soil surface temperature and light quality. Altered morphological traits due to agronomic decisions (row side or spacing) did not translate into higher seed yield, probably due to poor pollination and the wet conditions during harvest. Higher dry matter and leaf nitrogen was produced by ‘Legacy’ while ‘Quartz’ had more morphological structures (branches) to increase intercepted PAR. The environment of wider rows enabled the expression of these morphological characteristics for ‘Quartz’ and delayed critical leaf area which was favourable for flower development due to light quality. ‘Legacy’ had fewer branches and may produce higher yields when sown in narrower rows that achieve critical leaf area quickly to optimise PAR interception that can be converted to seed yield. Surprisingly leaf size did not differ between cultivars on the two dates it was sampled. However ‘Quartz’ exhibited many traits associated with smaller leaved cultivars while ‘Legacy’ responded as a larger leaved cultivar. ‘Legacy’ achieved critical leaf area quickly in narrower rows suggesting this cultivar is suitable for grazing systems however this poses challenges for seed production due to greater biomass production and subsequent shading from leaves.

New Zealand's farm systems rely on grazed pasture and fodder crops as the critical feed for livestock. 40% of New Zealand's agricultural land is in pastures which support livestock industries, generating an annual export revenue of NZ$33.8 billion. Due to climate change and tightening environmental regulations such as the 190 N application limit, the status quo perennial ryegrass white clover pasture is becoming less suitable for production in many regions of New Zealand, where water stress and waterlogging stress are increasing in frequency. Diverse pastures are a potential means to improve intra-annual yield stability and increase nitrogen fixation. To determine which species could be suitable for use in a diverse pasture mix for a farm conversion in Golden Bay, a randomised block experiment was set up to test 12 lines of perennial forages. The species included white clover, red clover, Alsike clover, Lotus pedunculatus, perennial ryegrass, cocksfoot, prairie grass, tall fescue and chicory. The species received treatments of waterlogging and drought stress. Drought plants were maintained just above PWP for 42 days, while waterlogging plants were submerged for 15 days, then removed and replaced in 3-day cycles according to stress levels. Under drought stress, all species had significantly reduced total biomass yields, ranging from a 55 % reduction in Alsike clover to a 25 % reduction in Caucasian clover. Under waterlogging, only white clover and red clover showed significant decreases in aboveground dry matter, but most species had at least an 80 % increase in senescence. Other traits measured included plant dry matter, photosynthesis, stomatal conductance, water potential, leaf senescence, leaf damage, SPAD and soil moisture. It was found that, on average, all other grass species were more tolerant in most traits than perennial ryegrass under drought, and all fescue lines were more tolerant than ryegrass under waterlogging. Compared with white clover, red clover and L. pedunculatus were more tolerant to drought and waterlogging, and Alsike clover performed similarly. The findings suggest suggest merit for examining diverse pastures based on L. pedunculatus, white cover, red clover, Alsike clover, chicory, cocksfoot, Continental and mediterranean fescues in the Golden Bay climate on Pakihi podzol. Among the legumes, L. pedunculatus was the best performing species for environments that experience drought as well as waterlogging. Despite similar biomass to white clover and red clover under drought and higher productivity under waterlogging, lotus had lower levels of senescence under both stresses, retaining greater root biomass under waterlogging. In grasses, hybrid tall fescue had the lowest levels of senescence under both stresses. Among the grass lines there were similar productivity levels and stress-induced changes. However, under both stresses, the fescue lines maintained the lowest levels of senescence among the grasses. Therefore, they could be tested as a suitable base to replace the traditional perennial ryegrass-based pasture used on Pakihi podzol in high rainfall environments. Future work under field conditions is needed to substantiate the findings from this research and progress the understanding of how the species interact with each other in response to the environment and grazing.

New Zealand dairy farms are no longer allowed to use high rates of nitrogen (N)-based fertiliser to achieve maximum pasture production due to the associated losses of N through nitrate leaching and nitrous oxide emissions. Gibberellic acid (GA), a naturally occuring phytohormone, has shown in short term studies, to increase, then reduce pasture yield when applied exogenously. The purpose of this research was to compare the long term effects of substituting N fertiliser with GA on a perennial ryegrass (Loliume perenne, PRG) and white clover (Trifolium repens, WC) pasture. A twelve-month, cut and carry plot trial explored dry matter (DM) production, botancial composition, ryegrass tiller size and density and forage nutritive value. To test these results, several treatment regimens were compared in a grazing trial with dairy cows which explored milk production and N excretion. Using a randomised block design with ten treatment regimens and four blocks, GA was applied to PRG and WC pasture plots to evaluate the timing and frequency of GA use in spring and autumn. Gibberellic acid replaced N fertiliser (once, twice or three times) in combinations of early (August and February), delayed (September and March) or late (October and April) applications. The control treatment was N fertiliser applied at 250 kg N/ha/y, the positive control was GA applied with every application of N fertiliser (80 g GA/ha/y + 250 kg N/ha/y) and the untreated control had no N fertiliser or GA. The first of ten applications occurred in August 2012 and the last in May 2013. Measurements included DM yield, botanical separation, tiller counts and nutrient content. Annual herbage yield was 14.6, 18.3 and 18.3t DM/ha/y (P0.05) on urinary-N content (0.4±0.02% N) or N excretion in milk (101±4.4 g N/cow/d). The lack of response to GA or N fertiliser highlighted the importance of timing of growth promotants for late lactation. In this study it was found that, the amount of N applied was more important for DM production than how much GA was applied. In spring delayed or late applications of GA, in place of N fertiliser was important to avoid N deficiency. While in autumn GA was able to be used in place of N fertiliser without affecting DM production. However growing times required in autumn may limit the effectiveness of GA on DM production and, consequently animal production.

Weed control is a critical component in pastures and hayfields in order to ensure maximum forage yields. Typically, broadleaf weed control in pastures and hayfields is achieved through the use of synthetic auxin. However, these herbicides also control desirable broadleaf species such as forage legumes, including white clover. Use of herbicides can lead to severe injury and often complete elimination of white clover, making it difficult for producers to maintain legumes in mixed grass-legume swards while controlling weeds. It is often desirable to have legumes present in the sward due to their high nutritive forage value and ability to fix nitrogen compared to grass only swards. Florpyrauxifen-benzyl + 2,4-D is a new herbicide which is reported to control broadleaf weed species, while preserving white clover. Little published research exists on this herbicide, particularly for use in pastures and hayfields. Research evaluating sward composition indicates that florpyrauxifen-benzyl + 2,4-D is effective in controlling broadleaf weed species while also preserving greater amounts of white clover than any other herbicide treatments. Florpyrauxifen-benzyl + 2,4-D also resulted in significantly more forage grass production than the nontreated control. Florpyrauxifen-benzyl + 2,4-D was less effective than other herbicides when applied via fertilizer impregnation. Additional research assessing the spectrum of broadleaf weed control found that florpyrauxifen-benzyl + 2,4-D is a viable herbicide for the control of several broadleaf weed species including bulbous buttercup, Canada thistle, broadleaf plantain, plumeless thistle, and common ragweed. However, florpyrauxifen-benzyl + 2,4-D was less effective than other herbicides for controlling certain weeds, such as horsenettle. White clover was injured from florpyrauxifen-benzyl + 2,4-D, but was able to fully recover in 90 to 120 days. There were no differences in white clover response between the four varieties tested. When evaluating establishment of forage species, florpyrauxifen-benzyl + 2,4-D did not injure or reduce biomass of tall fescue or orchardgrass plantings, indicating a high level of safety. Florpyrauxifen-benzyl + 2,4-D was also safe to both drilled and frost seeded clover when applied prior to and at planting. Greenhouse trials revealed that flowering white clover is more sensitive to herbicides compared to vegetative white clover, and that safety of white clover to florpyrauxifen-benzyl + 2,4-D is dependent upon use rate. Considerations such as weed species present, and the amount of white clover injury that is considered acceptable will dictate the decision to utilize florpyrauxifen-benzyl + 2,4-D in pastures and hayfields. This research demonstrates the effectiveness and overall utility of florpyrauxifen-benzyl + 2,4-D for use in pastures and hayfields due to the effectiveness of weed species as well as the level of safety to white clover.

The theory of Evolution by Natural Selection deals substantially with events now past, and with processes too slow for contemporary study. Experimental studies of evolution are possible, but are liable to the criticism that they deal with artificial, or at least extreme and atypical situations, involving rapid evolutionary change. Such criticism can be avoided to a degree by studying situations now in equilibrium. It is suggested that interesting results may be obtained by studying polymorphisms, and comparing the equilibria reached in populations of an organism in different environments. The chosen approach was to find a case of polymorphism convenient to study, followed by an attempt to interpret the pattern of its variation in terms of features of environments where a morph was common selecting in its favour, and vice versa. The material presented falls into three parts. 1) Material and Methods: Background and Preliminary Studies. The system chosen for study was the white V-shaped leaf markings of Trifolium repens. The species is widespread, abundant, and variable; the character is expressed on vegetative organs, and is therefore always accessible; and the known variation in this character had already been interpreted in terms of the action of a set of several alleles at one locus. The results of crosses indicate general agreement with previous reports. One heterozygous plant apparently largely failed to transmit one allele, Vb, through the pollen. Segregations indicate that a new phenotype, 'smeared', is determined by a particular Vh allele. Two more new phenotypes, 'shaded' and 'marginal' are described. The places where these and other rare phenotypes have been found are listed. New reports and the author's other experience are combined with previous reports in a new general account of leaf marking in the species, which also deals with at least three distinct classes of red markings. There appears to be more variation, implying more alleles, than previous accounts allow for, and reportedly distinct types, for example, Vh, Vl and Vf, seem to be linked by a series of intermediates, among which clear dividing lines are hard or impossible to draw. Experimental studies relating to the techniques of population sampling are reported. Scarification by 10 to 20 minutes exposure to concentrated sulphuric acid was chosen to deal with hard seed. The effect of a restricted number of mother plants on the accuracy of a seed sample is discussed; predictions concerning the frequencies of different types in progeny from single heads grown separately were roughly consistent with observations. Possible differences between vegetative and seed samples are discussed. Morph frequencies in a sample of plants bearing inflorescences were not found to be significantly different from those of the whole vegetative sample. The consequences of this species' ability to reproduce vegetatively are discussed, and a brief clone-mapping project is described. The procedures used to obtain and score material, and to obtain information about its background are described. The amounts of material of different types from different sources is summarised in Table 6.2. Problems involved in estimating gene frequencies are discussed, and the derivation of the values used to represent the different morph frequencies is outlined; they are basically phenotypic frequency figures, gene frequency calculations being used only where it was found to be necessary. 2) Observations on Population Samples. Polymorphism in respect of white leaf marks was found to be present in all except 15 out of 624 samples. The commonest group of phenotypes are the 'simple V-marks', referred to as 'L'. The next commonest is the unmarked type, 'O', present in all except 72 out of 624 samples, with an overall frequency among the plants scored of 17 per cent. Study of the sample data reveals a deficiency of the double marked phenotypes expected to be showed by VbyVl plants. This is explained as the result of a degree of dominance of Vby in such combinations; the effect of such dominance is allowed for in the frequency figures representing the frequency of 'L' marks, and of double-marks containing two members of the 'L-series' of simple-V-producing alleles. The possibility of demonstrating interaction between the frequencies of different morphs is discussed. There are indications of lower frequencies of 'By', 'B', and 'F' marks when the unmarked phenotype is common. The possibilities for the main object of the work, discovering associations between morph frequencies and environmental factors are shown in Table 8.1, giving the sets of data presentable for particular methods of analysis. Data on 148 British samples, scored in the field by the author, were treated by Multiple Regression Analysis (using the KDF 9 Computer of the Oxford Computing Laboratory), to test for association of morph frequencies with geographical location, altitude, and soil pH. Significant increases in the proportion of unmarked plants are shown with greater distance north and higher altitude. Frequency data for all the classes of morphs described is shown in the form of maps, of the British Isles, Western Europe, and the whole of the species' natural range. These maps confirm the northwards increase in frequency of the unmarked form found in Britain. Maps for the rare morphs show various patterns, most of which seem to involve central regions of higher frequency north of the Mediterranean (France - Alps - Greece) with lower frequencies elsewhere. Regression analysis of a set of seed samples from Spain supports the conclusion from British vegetative samples of an increase in unmarked frequency with higher altitude. Examination of the information about the background of other samples suggests higher unmarked frequencies in pastures than on waste ground, meadows being intermediate. British data suggest an association with wet, and particularly with badly-drained sites, with trodden paths, and with dense vegetation. There are patterns of response to water regime elsewhere but (e.g. Polish data) they tend to suggest the opposite association, of high unmarked frequency with dry conditions. 3) Comparison with other species. New observations are reported on some other related species. Available information on marking in other species of Trifolium is reviewed, and it is pointed out that it is scattered, difficult of access (much unpublished), and sometimes ambiguous or contradictory. Summary of this data indicates the presence of white V-markings in nearly 30 species of Trifolium, and its absence in at least as many more. Marks are probably entirely absent in subgenus Chronosemium and perhaps in part of subgenus Trifolium. but seem to occur in most of the other subgenera for which there is information. When marks are present, it appears that they are nearly always variable. Only three species are definitely reported as always marked, and in two of these there is variation between different types of marking. This suggests that the factors producing or preserving polymorphism in T. repens act also in other marked species. The presence of red leaf marks of various kinds is reported in 21 species of Trifolium. Also in the tribe Trifolieae, both red and white V-markings are found in Parochetus communis. Material grown by the author showed great variability within each plant in leaf mark, but no clear differences between plants. Red leaf marks, variable, and in some cases approaching a V-shape, are also found in Medicago. Several lines of evidence indicate relationships between the white markings, and the various red marking systems in T. repens. It is suggested that the white and red V-markings have a common evolutionary origin. Some examples of leaf marking with analogous properties in genera unrelated to T. repens are also briefly reviewed. In discussion some possible challenges to the validity of the results claimed are discussed, and evidence is presented suggesting that the reported genetic clines in unmarked frequency are real. Selective factors affecting unmarked frequency are tentatively suggested to be temperature and water regime. The problem of relating these to markings on leaves is discussed, and also the possibility that the phenomena of leaf marking are by-products of unknown processes, and are of no intrinsic importance. It is suggested that the inter- actions between red and white markings support the hypothesis that leaf markings themselves are of selective importance; and some possibilities as to what form this selective importance might take are mentioned. Possibilities for further work indicated in the course of the studies presented here are discussed. They include studies on important problems of population dynamics, which affect the design of techniques for sampling for leaf markings, but in which also observations on leaf markings could be used as means to ends of wider significance. Ways in which the methods used in the present study could be improved in a repeated study are suggested; however, it is felt that the clines observed provide a starting point for experimental work, and that this might be more rewarding than further descriptive work.

Trifolium is the most important pastoral legume genus for temperate agriculture. However, there has been little effort into characterising variation in Trifolium accessions. A series of studies utilising pedigree and genomic data were conducted to analyse variation in Trifolium accessions in the Margot Forde Germplasm Centre and in white clover breeding populations. Pedigree analysis experiments were designed to develop pedigree maps, calculate inbreeding and kinship coefficients, calculate the effective number of founders and identify influencing founders and ancestors for Trifolium repens (white clover), Trifolium pratense (red clover), Trifolium arvense, Trifolium ambiguum, Trifolium dubium, Trifolium hybridum, Trifolium medium, Trifolium subterraneum and Trifolium repens x Trifolium occidentale interspecific hybrids. One of the earliest parental white clover accessions identified was ‘North Canterbury Type I’ from 1941. The Type 1 phenotype influenced the population structure of all white clover accessions in the genebank. The relatedness and inbreeding coefficients revealed distinct germplasm pools formed across time that are of interest to pre-breeding efforts. One germplasm pool contained the majority of commercial white clover cultivars, of which they had similar phenotypes. Inbreeding in red clover remained stable in the last three decades, and a relationship between inbreeding and new introductions into the collection was found (r = 0.62). Founding accessions were classed into three phenotype groups; English Broad, English Giant Hybrid and Cotswald Broad. The first synthetic form was identified from parental accessions from English Broad and English Giant Hybrid. Within the minor Trifolium species, kinship levels remained below 8%, and no inbreeding was found. T. ambiguum and T. medium had the highest cumulative kinship across the decades. The Australian cultivar ‘Monaro’ had a strong influence over all of the studied accessions in T. ambiguum. Two genome-wide associations studies were designed to test for phenotype-genotype associations. The Mainstay panel had 242 white clover half-sib families, and the Genetic Gain panel had 80 white clover cultivars. The Mainstay panel had stolon density and growth score phenotype data. In contrast, the Genetic Gain panel had yield, clover content, seed yield, flowering duration, the peak number of flowers, leaf size, seed per head, leaf marking, cyanogenesis, normalised transpiration rate and the fraction of transpirable soil water phenotype data. Population structure was identified in the Mainstay panel (k=2), but no subpopulation divergence was present in the Genetic Gain panel. Both panels had rapid linkage disequilibrium decay. A marker-trait association was identified in the Genetic Gain panel for cyanogenesis. However, when the marker was tested against the marker dosage, it was not significant. The results of the genome-wide association studies emphasise the highly variable and heterozygous nature of white clover populations and cultivars.

The objectives of this research were to determine the impact of supplementing pasture-based diets with forage crops, previously used to mitigate winter N losses, on dry matter intake (DMI), milk yield, milk composition, and N utilization from dairy cows. Three supplementation studies integrating fodder beet (FB, Beta vulgaris L) or oats (Avena sativa) with perennial ryegrass (Lolium perenne L.)-white-clover (Trifolium repens L.) pasture were carried out in early or late lactation. Further, modelling studies were conducted to compare the effects of these crops on the productivity and profitability of irrigated dairy farms in Canterbury, New Zealand. The aim of the first experiment was to determine the effect of FB or oats on dry matter intake (DMI), milk production and N utilisation of grazing dairy cows in early lactation. The experiment was a comparative study of four spring feeding regimes with sampling replication via animal (n = 12 cows/treatment). Forty eight early lactation dairy cows were fed 3 kg DM/cow/d of fodder beet, oats forage or oats silage as supplement + 18 kg DM/cow/d of pasture for 21 days using a completely randomized design. Total apparent DMI was greater (P < 0.001) for grazed oats forage (OF) and oats silage (OS) compared with pasture only (Control) or fodder beet (FB). Cows substituted pasture for supplement maintaining a similar metabolisable energy (ME) intake (226± 3) across treatments. Consequently, there was no effect (P = 0.865) of supplement on milk yield or milk solids yield (P = 0.436). However, supplementation with FB resulted in lower (P < 0.001) urinary N concentration (4.7 g N/L) compared with CON, OF and OS (5.9, 6.0 and 5.5g N/L) respectively. The higher urinary N concentration in CON, OF and OS is likely to be due to higher N intake (630, 725 and 657 g N/cow/day respectively) compared to FB (589 g N/cow/day). These results showed there was no increase in milk production due to high substitution rates and similar energy intake. However, low protein supplements such as fodder beet were effective at reducing N intake and improve N use efficiency for milk production. Questions remained regarding the animal response to FB or oats when supplemented to cows with a lower energy demand in late lactation. The aim of the second experiment (autumn) was to determine the effect of FB or silage (ryegrass or oat) supplementation on DMI, grazing behavior, milk production and urinary N excretion on grazing dairy cows in late lactation. Fifty-four late lactation dairy cows were supplemented 4 kg DM of fodder beet (FB), oats silage (OS) or ryegrass silage (RGS) in addition to allocation of 12 kg DM/cow/d above 1500 kg DM/ha residual as perennial ryegrass and white clover pasture. The experiment was carried out over 21 days. Total DM intake was similar across treatments (P = 0.69), but milk production for cows supplemented with OS was lowest (1.01 kg MS/cow/day) compared with the RGS (control) or FB treatment that had similar milk yield (1.11 and 1.13 kg MS/cow/day respectively). The lower milk yield of OS compared with other supplements could be explained by the low ME of the supplement as well as the greater pasture mass offered to those cows. Autumn is an important time of the year to reduce N losses and supplementing with FB in late lactation reduced (P < 0.05) apparent N intake, milk urea N, (37, 40 and 44 g/dL) and spot urine N concentration (2.8 versus 3.9 and 4.0 g/L) compared to feeding OS or RGS respectively. These results demonstrated that cows on pasture offered low CP, and high-energy supplements, such as FB, during late lactation can sustain milk production while lowering N surplus. Further considerations were needed to understand why no milk response to FB was observed and to determine whether the quality of the product (milk composition) alters even though the quantity (milk yield) does not. In the third experiment, the purpose was to more clearly understand the lack of response in the first experiment by including grazing behavior measurements as well as the the effect of milk quality. Thirty-six early lactation dairy cows were fed perennial ryegrass-white clover (control) and supplemented 3 kg DM of FB or OF for 21 days. Again, as with experiment one there was no difference observed in total DMI (16.1 ± 0.52 kg DM/cow/day) or milk solids yield (1.9 ± 0.8 kg MS/cow/d) between treatments. Cows in each treatment spent similar amount of time grazing (413 ± 38.5 minutes/cow/d). There were supplement effects on milk quality. Milk fat and lactose (%) increased by feeding FB. The increase in fat % appeared to be due to increased short chain fatty acids. Though FB reduced the more valued long chain, (and some medium chain) fatty acids compared with the RGS control. Milk protein % was unaffected by supplement, but feeding FB increased casein concentration along with the minerals calcium and phosphorus. This suggests that cows supplemented with FB in early lactation may have been Ca deficient. As with previous experiments, cows that were supplemented with FB had lower N intake and lower urinary N concentration (2.4 g N/L) compared with OF and CON (4.3 and 3.1 g N/L). This resulted in, FB cows having a higher N use efficiency (36.2%) compared with OF and CON (29.5 and 29.8%, respectively). Given that there were little apparent milk yield benefits to supplementing with FB or oats during these short term grazing studies, follow up questions arose regarding the economic value of trying to integrate winter crops on the milking platform. To determine the feasibility and profitability of including winter supplements on the milking platform, results from experiments 1,2 and 3 were incorporated into a commercial decision support tool, FARMAX. Four scenarios were compared to a baseline farm system representing a Canterbury dairy farm. The scenarios were 1. Feeding FB only in spring, 2. Feeding FB (grazed) in autumn and drilling oats after grazing FB, which is grazed in spring, 3. Feeding oat forage only in spring and the last scenario was feeding oat silage in autumn. The results of the simulation showed that FB & OF (Autumn & Spring) scenario had the highest milk production (460 kg MS/ha) compared to (422, 418, 440 and 436 kg MS/ha) for baseline, FB spring, OF and OS scenarios respectively. Also, FB & OF (Autumn & Spring) scenario had lower cost per kg MS ($3.7 kg MS). The most profitable system at a $6.00/kg MS was FB & OF (Autumn & Spring) scenario on the milking platform. Results showed that despite there being no difference in herbage diet quality, supplementing FB & OF (A & S), scenario increased total DM offered and consumed (6.2%), thereby increasing milk production with 8.3% compared to (-1.6, 4.1 and 3.2%) for FB spring, OF and OS scenarios respectively with FB spring scenario that had the lowest milk production. Resulting in an increase of 10.1% in the gross margin profit/ha for FB & OF scenario compared to (6.0 and 4.4%) for OF and OS scenarios with the FB spring scenario having the lowest. In the short term grazing studies, supplementing pasture-fed dairy cows with FB or oats did not increase total DMI in early or late lactation due to high pasture substitution rates (SR) and similar energy intake. However, supplementing FB had effects on milk quality in both spring and autumn experiments. Only FB supplement lowered N intake (in spring) and improved N use efficiency for milk production as well as consistently lowering spot urine N concentration. Oats was more likely to reduce milk production when offered ensiled and may be more appropriate as a dry cow supplement. In both short term feeding and economic modeling, integrating FB supplement has value in sustaining milk production while lowering N losses in urine

The use of nitrogen (N) fertilizer is generally restricted on mixed species forage systems due to its stimulatory effect on grasses which increases competition with legume species. Reduced legume growth from this competition can compromise forage nutritive value and prospective yields. The controlled-release nature of several enhanced efficiency fertilizer N products holds the potential to improve legume persistence in mixed species pastures while providing supplemental N required by the grass component. The studies contained in this dissertation evaluated the effect of different enhanced efficiency N formulations (ATU, ESN, methylene urea, SuperU, and a 75% ESN: 25% urea blend) and untreated urea on yield, nutritive value, and legume persistence in a ‘Wrangler’ bermudagrass and ‘Durana’ white clover mixture (2014-2016 growing seasons), ‘KY-31’tall fescue and ‘Kenland’ red clover mixture (2015-2016 growing seasons), and ‘KY-31’ stockpiled tall fescue (2015-2017). The three studies were conducted at the University of Kentucky Spindletop Research Farm in Lexington, KY in a randomized complete block design. In the bermudagrass-white clover study, all enhanced efficiency N sources maintained white clover populations similar to the unfertilized grass/clover control, but only ESN caused greater clover composition than standard urea. Total forage yields increased linearly with N rate in all years, but dry weather conditions in the second and third years resulted in lower total yield. Forage nutritive value followed general trends throughout each growing season, but ESN’s ability to maintain clover resulted in higher nutritive value. In the tall fescue-red clover, total forage yields curvilinearly increased with N rate in 2015 but did not vary in 2016. ESN and ESN+urea blend treatments retained clover composition similar to that of the unfertilized control. Stockpiled forage yield increased with higher N rates. Enhanced efficiency N fertilizers with the ability to control N release can enhance forage yield while maintaining clover in mixed species swards.

Pastoral agriculture in New Zealand is based around the use of white clover and perennial ryegrass. Diploid perennial ryegrass, white clover and plantain were grown in monocultures, two-species mixes and three-species mixes at Lincoln University, Canterbury New Zealand. The dry matter production, botanical composition during two harvests on the 4th of August and the 13th of September 2017 were measured. Ten seed mixes were repeated at two overall sowing rates equivalent to 1000 and 2000 seeds/m2 and at two overall levels of N, 0 and 100 kg N/ha, respectively. A statistical analysis was performed on the data that was obtained from the measurements and a special cubic model was created. This analysis showed that pasture mixes with two species showed better estimated yields than the three-species pasture mixes. Ryegrass and white clover mixes and white clover and plantain mixes showed significant results (P80% in the monocultures compared with ryegrass and plantain which out competed weeds, only producing a weed content between 5-20%. White clover did not perform as a monoculture with the weeds and soil temperature following an autumn sowing not allowing for good establishment. A three-species pasture mix showed a decrease in production compared with the two-species mixes (P

The New Zealand agricultural sector is challenged with finding approaches to reduce the loss of nitrogen (N) through nitrate leaching, without the loss of farm production and profitabilty. One way to do this is to use different forages, regrowth intervals and N fertiliser rates to manipulate herbage N concentration and plant nitrogen use efficiency (NUE). The objective of this thesis was to quantify, for a range of pasture forages, herbage dry matter (DM) yield, chemical composition, N solubility and herbage N fertiliser requirements in response to different rates of N fertiliser, and regrowth interval in irrigated pastures in Canterbury. The thesis reports on four experiments. The objective of experiment 1 was to quantify the effect of N fertiliser rate on herbage DM yield and chemical composition, of two herb species (plantain and chicory) three legume species (lucerne, red clover and white clover) and seven grass species (prairie grass, Italian ryegrass, tall fescue, tetraploid perennial ryegrass, diploid perennial ryegrass, high sugar perennial ryegrass and cocksfoot) at optimum defoliation time over a two year period. In this study, conducted on the Canterbury Plains, New Zealand, the twelve species were grown at six N fertiliser rates ranging from 0 to 450 kg N/ha/year and managed under irrigation and cutting management. Herbage DM yield and N concentration were measured at each harvest over two years (1 December 2014 to 30 November 2016). Defoliation management was on average, every 29 days in grasses and herbs with a total of 9 harvests per year, and 38 days in legumes with a total of 7 harvests per year. Forages were not harvested over the winter (June, July). As N fertiliser rate increased from 0 to 450 kg N/ha/year, annual herbage production increased linearly, from 5016 to 13812 kg DM/ha/year in grasses and from 6844 to 11481 kg DM/ha/year in herbs. In contrast, annual DM yield for legumes was unaffected by N fertiliser rate and ranged from 10596 to 10936 kg DM/ha. Additionally, there were contrasting (P3.2% N, averaged over the N fertiliser rates), compared to spring (2.5 % N, averaged over the N fertiliser rates) and increased with N fertiliser rate (3.1 N % at nil N fertiliser to 3.9 N % at high N fertiliser, P

An investigation of how nutritive value changes with respect to the number of species and proportions of species in pasture mixtures was carried out at Lincoln University. Tetraploid perennial ryegrass cv. Base, plantain cv. Tonic, white clover cv. Apex and red clover cv. Grasslands Sensation were all grown in monocultures, two species, three species and four species mixtures based on a simplex centroid mixture experiment design. The aim of the experiment was to quantify the nutritive value of each of these mixtures to be able to define an optimum pasture mixture based on the four species used. Allowing for the species contribution to nutritive value and diversity effects to be identified. Models have been produced to predict nutritive value attributes of a mixture involving the four species investigated. The optimal mixture was defined based on parameters of maximising annual dry matter yield and annual yields of crude protein and metabolisable energy while meeting target annual mean concentrations of metabolisable energy, acid detergent fibre, neutral detergent fibre and crude protein based on animal requirements. This mixture comprised of the following proportions: 0.43 ryegrass, 0.20 white clover and 0.37 red clover based on the number of seeds/m2. Which was equivalent to the sowing rate of 12.90 kg/ha of ryegrass, 1.50 kg/ha of white clover and 6.48 kg/ha of red clover which was equivalent to a total sowing rate of 20.88 kg/ha. This mixture would allow for the pasture to meet animal requirements while not limiting intake or stocking rate. The optimal pasture mix was expected to have an annual yield of 14,250 kg DM/ha, annual metabolisable energy yield of 160 GJ/ha and an annual crude protein yield of 2806 kg/ha. It was also expected to have metabolisable energy concentration of 11.2 MJ ME/kg DM, an acid detergent fibre concentration of 264 g/kg DM, a neutral detergent fibre concentration of 380 g/kg DM and a crude protein concentration of 195 g/kg DM. The optimal pasture mixture would allow for protein and energy requirements of livestock to be met without limiting intake. Nutritive value was found to be dependent on the species present and their proportions in each mixture. Nutritive value has shown variation across years and seasons which showed that diversity effects among species were not consistent across time. The dynamic systems of pasture mixtures show that proportions of species do not remain consistent across time. Therefore, the diversity effects were not consistent across time either. Proportions of species did not remain consistent with the sown proportions with white clover being most outcompeted within the sward often stabilised at 10% of the mixture. The proportions of ryegrass, plantain and red clover varied depending on the mixture. The nutritive value was subsequently remodelled based on the actual proportions of species in the mixtures. In conclusion, the nutritive value of the pasture mixtures were dependent on the species present and the proportions of each species. The proportions of species changed across seasons due to growth patterns and overs years as a result of the succession of species.

Nitrogen (N) leaching from soil is a significant concern for intensively grazed forage-based systems because it can cause a decline in water quality. Urine patches are the main source of N leaching in these systems. In New Zealand, increasing pressure to increase export earnings, while reducing N leaching loss from agriculture, poses a challenge for farmers and mitigation options are urgently needed. One approach is to increase the uptake of urine-N by forage plants, thereby reducing N leaching. The aim of this PhD programme was to increase our knowledge and understanding of the effects of alternative forages and gibberellic acid (GA) on N leaching from grazed agricultural soil. Three lysimeter experiments were conducted. The first quantified the effect of forage type (perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.) (RGWC), Italian ryegrass (Lolium multiflorum Lam.), and lucerne (Medicago sativa L.)) and GA application on N leaching, herbage dry matter (DM) yield, and N uptake. Lysimeters (0.5 m diam., 0.7 m deep) were treated with urine (700 kg N ha⁻¹ ), and ± GA (8 g GA ha⁻¹) in May 2014. A ¹⁵N balance was conducted to determine the fate of the applied urine-N. The second lysimeter experiment determined the N leaching loss, herbage DM yield, and N uptake from the urine patch of an Italian ryegrass, plantain (Plantago lanceolata L.), white clover mixture (Italian-Plantain Mix), compared with RGWC. Urine was applied in March 2015 and included a 700 kg N ha⁻¹ (Urine 700) treatment, and urine-N at the concentration excreted by cows grazing each forage type (Urine Actual). The third experiment used smaller lysimeters (0.18 m diam., 0.5 m deep) to determine the N leaching loss, herbage DM yield, and N uptake response of RGWC to an application of GA (8 g GA ha⁻¹) over a range of urine-N rates (0, 25, 50, 100, 200, 400, and 700 kg N ha-1) applied in April 2016. A pot experiment was also conducted to determine whether there were any differences in the soil ammonia-oxidising bacteria and archaea beneath perennial ryegrass and Italian ryegrass, these were compared with bare soil. Pots (0.0144 m², 0.13 m deep) were destructively harvested 1, 15, 30, 61, and 90 days following urine application (700 kg N ha⁻¹) in May 2015. Significant reductions in N leaching loss, were shown for Italian ryegrass (35%) and Italian-Plantain Mix (45-89%) forages, when compared with RGWC. The mechanisms behind this were a reduction in urine-N excretion (for Italian-Plantain Mix), and increased cool-season uptake of urine-N by the Italian ryegrass. This was reinforced when no difference in soil ammonia-oxidisers was shown between perennial ryegrass and Italian ryegrass which suggested that Italian ryegrass was not inhibiting nitrification. Gibberellic acid had no effect on N leaching. This research has clearly shown Italian ryegrass, and Italian-Plantain Mix as promising alternatives to RGWC, which could reduce N leaching losses from grazed systems. Lucerne is not recommended as an alternative to RGWC, as N leaching losses were higher under grazed conditions. Gibberellic acid had no direct effect on N leaching loss and so is not recommended as a direct mitigation tool for N leaching losses in grazed systems.

Expansion and intensification of dairy in pasture based systems whilst economic, creates environmental challenges through nitrate leaching and degradation of water quality. With pressure to reduce nitrate leaching and simultaneously increase pasture productivity and milk production, studies have been carried out to develop management strategies to reduce N surpluses. At the animal level an effective way to reduce the environmental impacts is through nutritional interventions. Inclusion of plantain (Plantago lanceolata L.), in the traditional perennial ryegrass-white clover pastures has shown potential to reduce nitrogen concentration in urine. Most studies including plantain however, grew it in diverse pastures where it is difficult to manage persistence due to different grazing requirements compared with perennial ryegrass. Further its contribution to N loss when sown in a mix with other plant species was not clear. This research evaluated the effect of grazing increasing proportions of spatially planted plantain and perennial ryegrass-white clover pastures species on N excretion, milk production, urinary and feeding behaviour in Canterbury, New Zealand. The experiment was carried out at Lincoln University Research Dairy Farm, Canterbury, New Zealand in April 2016. Forty eight late lactating primi and multi-parous Friesian x Jersey cows were blocked according to pre-experimental average (mean, ± s.e.m) liveweight (499.21 ± 6.44 kg), milk solids (1.41 ± 0.02 kg), milk yield (14.14 ± 0.29 kg) and age (5.42 ± 0.19). Four cows were randomly allocated to 3 replicates of 4 treatments (0% plantain, 15% plantain, 30% plantain, 60% plantain) of proportions of spatially separated perennial ryegrass-white clover and pure plantain. Pre-grazing herbage mass (kg/ha DM ± sem) for pasture and plantain were similar across treatments (3857.9 ± 46.5 and 4973.1 ± 74.5 respectively) whilst the cows grazed the two monocultures to the same post-grazing herbage mass (1504.8 ± 36.1 vs. 1510.5 ± 79.3). Chemical composition was similar between perennial ryegrass-white clover and plantain across treatments. Pre-grazing pasture had similar CP and ME (23.6 ± 0.61%, 12.3 ± 0.03 ME/kg DM and 20.5 ± 0.24% and 12.77 ± 0.04 ME/kg DM for perennial ryegrass-white clover and plantain respectively. Dry matter intake (DMI) estimated from pre- and post-grazing herbage mass was lower (P=0.003) for PL0 than PL15, PL30 and PL60 and similar between PL15, PL30 and PL60 (14.74 vs. 15.96, 16.00, 15.97kg respectively). Milk yield (16.1 L/cow/day ), milk solids (1.6 L/cow/day), protein ( 4.5%) and fat (5.8%) were unaffected by percent proportion increase of plantain in the diet. Urine N concentration and urine N output had 33% lower N (P=0.012) in the PL60 than PL0 (0.30 vs. 0.45 g N/L and 431.3 g N /day vs. 545.4 g N/day respectively). The effect was mainly related to the difference derived from 30% to 60% plantain in the diet with the difference in N concentration and N output between PL0 and PL30 being 13%. Milk, urea and milk urea N declined (P=0.004) by 18% in PL60 compared with PL0. The urine volume (2.3 litres/urination) and patch area (0.34m2) were unaffected by the percent increase of plantain proportions in the diet. Based on urine volume, patch size and N concentration, there was a 40% decline from 316 to 190 kg N/ha in urine patch loading from 0% to 60% proportion of plantain in the diet indicating the key potential of plantain to reduce N loading in pastures. As urine loading is critical for determining N leaching, this is likely to contribute to reduced leaching even if urine frequency is marginally higher. The cows allocated more time (47.3 minutes) to plantain than pasture (43.54 minutes) during the morning grazing, but in the afternoon grazing bout, the reverse was true with more time spent grazing pasture than plantain (62.1 vs. 34.9 minutes/cow respectively). Animal bite rates were similar in the morning and afternoon grazing bouts (46 vs. 39 and 49 vs. 44 bites/minute/cow respectively). This study confirmed that plantain has similar feed value and milk production potential to perennial ryegrass-white clover and when offered as green leafy herbage to dairy cows. However, feeding greater than 30% plantain in the diet resulted in significant reduction in both urine N concentration and urine N excretion. As these are key components of determining N loading of urine patches, the study indicates that plantain may present appealing opportunities to reduce environmental impact of dairy farming.

Successful establishment of white clover (Trifolium repens) and lucerne (Medicago sativa) in the field depends on the formation of effective symbioses with Rhizobium leguminosarum and Sinorhizobium meliloti, respectively. High numbers of naturalised rhizobia and the poor persistence of commercial inoculants has cast doubt on the need for commercial inoculation in New Zealand. This study investigated aspects of the soil ecology of R. leguminosarum and S. meliloti that affected the persistence and success of applied strains. In planta the commercial inoculant, strain S. meliloti strain RRI128, was present in 57-61% of lucerne nodules four years after sowing when applied as a peat slurry. In contrast, R. leguminosarum strain TA1 was only present in 8% of white clover nodules eight months after sowing. Naturalised strains of rhizobia were also present in the nodules of both species. The abundance of naturalised strains differed seasonally, from 5.2 x 104 cells g-1 soil in the winter to 200 cells g-1 soil in the summer. This was explained by increasing numbers at low (20%) soil moisture. Additional bacteria, including a Pseudomonas sp. capable of forming nodules, were also isolated from both plant species. Illumina sequencing revealed that Rhizobium and Sinorhizobium were the dominant genera in the nodules of white clover and lucerne plants, respectively, regardless of inoculation treatment and made up 89-97% of all sequences. Sequencing confirmed the presence of other bacteria in the nodules of both host species with 408-879 genera identified, with each represented by 1.5) of many carbon sources and was more competitive for nodule occupancy compared with TA1, which utilised fewer carbon sources at lower amounts (average absorbance: 70% plants nodulated) at pH 5.1 compared with RRI128 (100 vs.

Nitrogen (N) from urine patches is a major contributor to N leaching, due to the high loading rate of N in urine patches compared with the capacity of many plant species to take up N. This study aimed to determine milk production and urinary N excretion from dairy cows grazing a single alternative forage; plantain; to compare the effect of time of day and N fertiliser inputs on the chemical composition of alternative forages which may prove useful in reducing the supply of N to the grazing animal and; to investigate management strategies around time of day allocation which allow for at least 50% of the animals diet to be plantain. Five experiments were conducted in Canterbury, New Zealand which compared forages in grazing, small plot and in sacco studies. In the first experiment, the effects of 50% or 100% of a herbage diet of plantain on milk production and urinary N concentration were measured in two experiments for late (autumn 2015) and early (spring 2015) lactation spring calving cows. Three groups of 12 mixed age Friesian × Jersey dairy cows were offered a perennial ryegrass-white clover pasture, pure plantain or 50% perennial ryegrass-white clover and 50% pure plantain by ground area (50–50 pasture–plantain). Cows were blocked according to milksolids, age, days in milk and liveweight. The experiment was conducted for 10 days with a 3 day transition period and 7 day experimental period. Urine N concentration was lower in the late and early lactation experiments (P 0.05) between treatments. Under the conditions of the experiment, allocating plantain temporally resulted in lower milk production. When offering a diet of 50% plantain - 50% perennial ryegrass, this experiment suggested allocation for greatest milk production is best achieved through spatial separation. A feature of the results from the grazing experiment of this study was the reduction in urinary N concentration when cows grazed plantain, despite similar N intake to cows grazing perennial ryegrass-white clover. Other research suggests this may be due to plant secondary metabolites aucubin, acteoside and catalpol. A final experiment aimed to characterise the concentration of secondary metabolites in plantain herbage. This was a two year experiment using plantain from the plot experiment. The experiment was a randomised block design with three replicated treatments. Treatments were two N fertiliser rates (180 kg N/ha/year and 450 kg N/ha/year) and two harvest times (0700 h and 1600 h). Samples were taken seasonally when plantain had reached grazing height. Concentrations of the metabolites were varied. Acteoside consistently had the greatest concentration (ranging from 36.0-1.77 mg/g DM), aucubin intermediate (range 5.85-0.566 mg/g DM) and catalpol lowest (

A pasture mixture experiment was conducted, growing four species; perennial ryegrass, plantain, white clover and red clover under irrigation in Canterbury, New Zealand. The experiment set out to identifty an optimal pasture seed mixture that maximised annual dry matter (DM) yield by quantifying species identity and diversity effects from pasture mixtures. The plots were sown containing one- to four-species, based on a simplex mixture design containing all possible combinations. The annual and seasonal DM production were statistically analysed with a special cubic model created from the simplex mixture design, enabling predictions to be made on the mixture that would produce maximum yield for the second year of growth (May 2016 to May 2017). The DM production and botanical composition of each monoculture and mixture were measured, compared and analysed. To explain differences in yield between mixtures, N content, leaf area index and light interception were also measured. The modelled analysis showed that five two-species mixtures and one three-species mixture showed significant species interactions and subsequent yield increases over the mean of their component monocultures. The seed mix that was predicted to maximise yield consisted of 0.29 ryegrass, 0.20 plantain and 0.51 red clover in terms of seeds/unit area, predicted to yield 16.44 t DM/ha for the second year of growth. This mix produced 5.19 t DM/ha more than the rygrass and white clover (RG*WC) mixture, which yielded 11.25 t DM/ha. The greatest diversity effects were shown by the two species mixtures that included red clover, and produced 4.93 t DM/ha (RG*RC) and 5.31 t DM/ha (P*RC) above the mean yield of the monocultures. Species interactions and botanical compositions within the mixtures were not consisitent over time; the optimal mixture (ryegrass, plantain and red clover) displayed transgressive overyielding over six of the eight sampling dates. There were no significant species interactions across any of the mixtures at the early spring harvest (August 2016) and the late autumn harvest (May 2017). The botanical composition of the highest yielding mixture RG*P*RC vaired in botanical composition over the season. It deviated from its sown proportions of 0.33 RG -0.33 P - 0.33 RC to 0.24-0.54-0.22 of RG, P, RC respectively at the time of first harvest (2/08/2016). By the second harvest (22/09/2016) the botanical composition of the mixture had changed to 0.10-0.26-0.64; halving the proportion of ryegrass and plantain and tripling the proportion of red clover. There were significant differences in N content between the four monocultures, ranging from 2.30% (ryegrass) to 3.75% (red clover). The mixture N% was lower when white clover was included in the mixture (2.71% for RG*WC, 2.82% for P*WC) than when red clover was included (3.22% for RG*RC, 3.49% for P*RC). The N content of the mixtures can be used to explain yield as the average N concentration of the mixtures increased by 1%, the annual yield increased by 6.60 t DM/ha. Leaf area index and light interception analysis within the seventh and eighth harvests showed that significantly higher yielding mixtures such as RG*RC, P*RC and RG*P*RC reached critical light interceptance and had a significantly higher leaf area index than their component monocultures and mixtures that produced lower yields, such as RG*P, P*WC and WC*RC

The main objective of this thesis was to evaluate perennial ryegrass (Lolium perenne L.) cultivars for dairy production under different pasture management. Five experiments were conducted in Canterbury, New Zealand to evaluate perennial ryegrass in terms of morphology (experiment 1), nutritive value (experiments 1 and 2), rumen degradation (experiment 3), dietary preference (experiment 4) and milk production (experiment 5). Pasture management included growing perennial ryegrass cultivars with and without white clover (Trifolium repens L.) (experiments 1, 2, 3 and 4) and defoliating at different times of day (experiments 2 and 5) during pasture regrowth (experiment 2). In experiment 1, morphology and nutritive value of eight perennial ryegrass cultivars (AberMagic, Alto, Base, Bealey, Commando, Kamo, One50 and Prospect) growing with and without white clover were evaluated in a split-plot design at three phenological stages (pre-heading vegetative, reproductive and post-flowering vegetative). Pastures were managed under dairy grazing with irrigation and the annual nitrogen (N) fertiliser application rate was 325 kg N/ha. Compared with diploid cultivars, the tetraploid cultivars, Base and Bealey, had less dead material proportion (16.0% vs. 18.0% at the pre-heading vegetative stage; 22.0% vs. 27.0% at post-flowering vegetative stage) and greater lamina proportion (51.7% vs. 45.1% at the pre-heading vegetative stage; 55.0% vs. 50.7% at post-flowering vegetative stage) at the vegetative stages. The high-sugar diploid cultivar, AberMagic, had the greatest water-soluble carbohydrates (WSC) concentration at all stages in both lamina and pseudostem. Organic matter digestibility in dry matter (DOMD) was greater in Base, Bealey and AberMagic (mean = 75.9%) than other cultivars (mean = 73.8%). However, the differences in morphology was not the root cause of the variation in chemical composition and digestibility among cultivars. White clover had little effect on morphology of perennial ryegrass but significantly increased the crude protein (CP) concentration of accompanying perennial ryegrass at the pre-heading vegetative stage from 162 to 170 g/kg dry matter (DM). Interactions between perennial ryegrass cultivar and the presence of white clover were found for herbage WSC, acid detergent fibre (ADF) and neutral detergent fibre (NDF) concentrations and DOMD at the reproductive stage only. In experiment 2, herbage nutritive value variation of three perennial ryegrass cultivars (AberMagic, Bealey and Prospect) harvested at different times of day (from sunrise to sunset) during pasture regrowth (from 7 to 28 days after defoliation) was evaluated in monocultures and mixtures with white clover. AberMagic had a greater WSC concentration (162 g/kg DM) than Bealey (144 g/kg DM) and Prospect (112 g/kg DM). White clover proportions in perennial ryegrass swards (4 cm above ground level) increased during pasture regrowth from 6.5% on day 7 to 13.8% on day 28. The presence of white clover led to a greater overall herbage CP concentration (260 vs. 286 g/kg DM) and a lower NDF concentration (421 vs. 397 g/kg DM) in all cultivars. Herbage WSC concentration increased from 87 to 186 g/kg DM during the 4-week regrowth and was greater in the herbage harvested at sunset (180 g/kg DM) than sunrise (93 g/kg DM). In contrast, herbage CP concentration decreased during pasture regrowth from 292 g/kg DM to 242 g/kg DM and was lower in the herbage harvested at sunset (263 g/kg DM) than sunrise (293 g/kg DM). Herbage NDF concentration increased during pasture regrowth from 394 g/kg DM to 436 g/kg DM but dropped during the day from 439 g/kg DM to 378 g/kg DM. DOMD showed an opposite pattern to the NDF concentration, decreasing during pasture regrowth from 81.4% to 75.1% and increasing during the day from 76.4% to 81.2%. The interaction between perennial ryegrass cultivar and growth stage indicated that the cultivar ranking by CP and NDF concentration and DOMD changed during pasture regrowth and the differences in the WSC concentration among perennial ryegrass cultivars increased as regrowth proceeded. In experiment 3, four rumen-fistulated Friesian × Jersey dairy cows were used in two Latin square designs to investigate in sacco rumen degradation characteristics of perennial ryegrass cultivars in monocultures and in mixtures with white clover at different proportions. Perennial ryegrass cultivars included AberMagic, Bealey and Prospect, and white clover proportions in the mixtures were 0%, 16.7%, 33.3% and 100% on a fresh weight basis. The tetraploid cultivar, Bealey, had a greater soluble fraction a and a faster insoluble-degradable fraction b of both DM (0.305 and 0.621/h, respectively) and organic matter (OM, 0.280 and 0.647/h, respectively). Regarding CP degradation, Prospect had a greater soluble fraction, a (0.132), than AberMagic (0.082) and Bealey (0.062). Bealey had the lowest accumulative and instantaneous N release rate relative to OM disappearance, indicating it had the best rumen N/energy balance and synchrony among cultivars. As white clover had a greater effective degradability of CP (76.6% vs. 60.5%) and effective rumen degradable protein (ERDP) concentration than perennial ryegrass (237 vs. 118 g/kg DM), the effective degradability and ERDP concentration of the mixtures increased proportionally as white clover proportion increased. Consequently, white clover exerted a negative effect on rumen N/energy balance and synchrony. However, the impact was limited when the white clover proportion was low (16.7% on a fresh weight basis). In the mixtures, white clover interacted with perennial ryegrass for DM and OM predicted potential rumen degradability but did not the effective degradabilities. In experiment 4, dietary preference and selection of dairy cows for eight perennial ryegrass cultivars (AberMagic, Alto, Base, Bealey, Commando, Kamo, One50 and Prospect) growing with and without white clover was examined in a split-plot design at three phenological stages (pre-heading vegetative, reproductive and post-flowering vegetative). The dietary preference was defined as the decreasing rate of sward surface height at the beginning of grazing with minimum environmental constrains. The selection index was calculated based on pre-grazing herbage mass availability and post-grazing herbage mass residual of each cultivar. Results showed that the tetraploid perennial ryegrass cultivars, Base and Bealey, and the high-sugar diploid cultivar, AberMagic, were most preferred and selected by dairy cows. The proportion of perennial ryegrass, herbage WSC concentration and DOMD were positively correlated with dietary preference and selection indices, while negative correlations were found with the proportion of dead material and herbage ADF and NDF concentration. Although white clover had limited effect on sward structure and morphology of perennial ryegrass pastures, its presence (4.9% to 6.8% on a DM basis above ground level) increased overall herbage CP concentration (154 vs. 171 g/kg DM at pre-heading vegetative stage; 230 vs. 249 g/kg DM at post-flowering vegetative stage). The interaction between the presence of white clover and perennial ryegrass cultivar for dietary preference index indicated that the presence of white clover reduced the differences in dietary preference for perennial ryegrass cultivars. However, it did not lead to a significant re-ranking of cultivars preference order. In experiment 5, the effect of perennial ryegrass cultivar (AberMagic and Prospect) and the timing of herbage allocation (in the morning at 0830 h or afternoon at 1630h ) on herbage nutritive value and milk production of mid-lactation dairy cows were examined in a randomised block design. Twelve groups of four Friesian × Jersey cows were allocated to three replicates of four treatments over 10 days. There were no significant differences in sward structure and morphological characteristics between cultivars, except Prospect having a lower average tiller mass (43.1 mg DM/tiller) than AberMagic (48.4 mg DM/tiller). The concentration of WSC and DOMD was greater in AberMagic (180 g/kg DM, 74.2%) than Prospect (153 g/kg DM, 71.4%). Herbage DM percentage, WSC concentration and DOMD were lower in the morning compared with in the afternoon (22.3% vs. 18.8%; 154 vs. 179 g/kg DM; 72.1% vs. 73.5%). Herbage DM intake (mean = 12.0 kg/cow/day), milk yield (mean = 17.2 kg/cow/day) and milksolids yield (mean = 1.60 kg/cow/day) did not differ significantly among treatments. Milk urea nitrogen (MUN) concentration decreased when Prospect was allocated in the afternoon compared with in the morning (8.39 vs. 9.33 mmol/L), while MUN concentration remained similar whether AberMagic was allocated in the morning or afternoon (8.95 vs. 8.96 mmol/L). This interaction suggested that the effects of allocation timing may potentially be different for different perennial ryegrass cultivars. Pasture management exerted more pronounced and consistent effects than the effects of perennial ryegrass cultivar on herbage nutritive value. Interactions were found between perennial ryegrass cultivar and mangement for herbage nutritive value (the regrowth stage) and dietary preference (the presence of white clover). Therefore, pasture management could potentially be a more effective way to manipulate herbage nutritive value than perennial ryegrass cultivar. Additionally, the evaluations of perennial ryegrass cultivars should be conducted under a defined and consistent condition to exclude the management effects and cultivar × management interactions for herbage nutritive value. Further, in order to achieve a higher herbage feeding value, management could be adjusted to benefit a particular cultivar, and cultivars, in turn, could be selected to fit a certain management system better.

In forage-based livestock production systems, dietary N is often in excess supply. In addition, the efficiency with which ruminants utilise dietary N utilisation is low, leading to substantial losses of N to the environment, particularly in urine. There are considerable political and societal pressures for farmers to reduce N losses, yet still maintain productivity. To enable reductions in N excretion and improve N utilisation, the flow of N through ruminants needs to be understood. While much is known about the ruminal degradation of dietary N, less is known about the release of N from herbage during ingestion. As herbage N is readily fermentable, the extent of release of N during ingestion has implications for ruminal ammonia concentration and concomitantly, N excretion. The aims of this PhD research programme were to quantify the variation between forages in N release during comminution of herbage, explore the characteristics of herbage that influence N release, and consider implications for N excretion. Two laboratory based experiments were conducted to examine the release of N during mechanical maceration of herbage, as a proxy for the effect of ingestive mastication on herbage by dairy cows. The first compared N release from herbage of five commonly used forages: perennial ryegrass (Lolium perenne L.), white clover (Trifolium repens L.), lucerne (Medicago sativa L.), chicory (Cichorium intybus L.), and plantain (Plantago lanceolata L.), collected in February, May, August, and November of 2015. The amount of N (as crude protein, CP) released during maceration of herbage varied 15-fold, with lucerne releasing the most (11.5 g/100 g DM) and plantain the least (0.8 g/100 g DM). Crude protein release from forages ranked as follows: lucerne > ryegrass = clover > chicory > plantain. Investigations revealed that no single characteristic (i.e. herbage physical properties or chemical composition), measured in this study, was strongly associated with CP release of all forages evaluated. However, forages that tended to have greater herbage CP and non-protein N (NPN) concentrations tended to release more N during maceration. This finding led to the second laboratory experiment that explored the effect of increasing rates of N fertiliser application (0, 100, 200 and 350 kg N/ha/y) on herbage characteristics and release of N. Herbage was collected from swards of perennial ryegrass, lucerne, chicory and plantain in May (autumn) and November (spring) 2015. Increasing N fertiliser rate applied increased leaf length (up to 140%), herbage mass (up to 320%), and CP and NPN concentration in perennial ryegrass, chicory and plantain. Increasing N fertiliser rate doubled the amount of CP released from ryegrass herbage across both seasons, and chicory in autumn, but increased CP release from plantain 6-fold in autumn. Increasing N fertiliser rate applied to ryegrass, chicory and plantain swards also reduced the ratio of fermentable carbohydrate to CP in herbage, which is thought to increase loss of N in urine. The greatest increase in the amount of CP released and greatest decline in the ratio was observed at N rates exceeding 200 kg N/ha/y, coinciding with reductions in herbage mass and leaf growth response to increased rates N fertiliser application. This study suggests rates equivalent to 200 kg N/ha/y ought to be the upper limit of N fertiliser applied to swards of ryegrass, chicory and plantain. The third experiment conducted sought to determine the effect of ingestive mastication by mature dairy cows on fresh cut herbage from three forages (perennial ryegrass, lucerne and chicory). The ranking of herbages for CP release was similar to that determined in the first two laboratory based experiments, where 7.5, 5 and 3 g CP/100 g DM was released from lucerne, perennial ryegrass and chicory, respectively, during ingestive mastication. Release of CP was associated with the concentration of CP in herbage (R2 = 0.53) supporting the findings of the previous laboratory experiments. When N intake rate was calculated, delivery of N to the rumen was greatest in cows fed ryegrass (0.79 g N/min) compared with lucerne and chicory (0.54 and 0.22 g N/min), which would have implications for microbial utilisation in the rumen. Despite the slower intake rate of chicory compared to lucerne and ryegrass (46.5 vs. 64.7 and 70.7 g DM/min) the extent of comminution was less in chicory compared to the other forages (36 vs 44 and 41% of particles reduced to < 4 mm), but the difference was not associated with any of the herbage characteristics measured in this study, nor was comminution shown to be associated with CP release when data from the three forages were combined for analysis. This research programme concluded with a simulation study exploring milk-solids (MS) production and urinary N excretion from cows grazing swards of perennial ryegrass, lucerne, chicory and plantain grown under increasing rates of N fertiliser (0, 100, 200, 350 and 500 kg N/ha/y) using the data collected in the second experiment. Predictions of MS production and N excretion by the MINDY model for cows grazing chicory, ryegrass or plantain, suggests that rates of N fertiliser of 200 kg/ha/y could be the optimal point for MS production before total daily N excretion increases substantially. As at this application rate, DMI and MS production predicted for cows grazing chicory was near maximum (15.9 kg DM/cow/d; 1.24 kg MS/cow/d) with little gain achieved through additional N fertiliser but a 100 g N/cow/d increase in urine N excretion (42% increase) was predicted as N applied increased from 200 – 500 kg N/ha/y. In the case of plantain and ryegrass, while MS production was predicted to increase beyond the 200 kg N/ha/y fertiliser rate, the amount of N excreted also increased by 65 and 103 g N/cow/d, respectively (50 and 64%). The simulations also suggest that the addition of N fertiliser to lucerne swards has little effect on MS but increases urine N excretion, and is thus not recommended. Overall, the findings of this research program confirm the hypothesis that the release of N from herbage during comminution differs between forages. This finding can aid explanation of literature that has demonstrated differences in the N excretion by cows fed diets of different forages, and can help understanding of N flow in ruminants. The study also suggests that in autumn, there is potential to manipulate N release from perennial ryegrass, chicory and plantain herbage during maceration and to reduce N excretion from cows grazing diets of these forages through altering the amount of N applied to these swards. This information could be used to inform fertiliser management practices, and development of diets for dairy cows to reduce urinary N loss.

Nitrogen (N) and phosphorus (P) are limiting nutrients for plant growth and are the largest fertilizer inputs in agriculture. Phosphorus fertilizer has a low efficiency of use with only 5-30% utilized by the plant, the rest forming metal complexes and becoming unavailable to plants. In New Zealand, most pastures contain Trifolium repens (white clover) and T. subterraneum (subterranean clover) grown in mixture with forage grasses. Clovers are nodulated with Rhizobium spp. to fix nitrogen. These bacteria can also release organic acids, phytases and phosphatases to make P soluble and plant available. The goal of this thesis was to understand the potential of P solubilizing rhizobia (PSR) to perform dual functions of P solubilization (PS) and N fixation and determine the magnitude of growth benefit to white clover. To achieve this, the study was divided into four objectives: 1) to assess the populations of P solubilizing bacteria inhabiting clover nodule, 2) to profile the nodule bacterial community in different clover hosts and study how its affected by different soil P environments, 3) to increase the mechanistic understanding of PS and assess N fixation by PSR in vitro and 4) to assess growth promotion of clover by PSR. A total of 1,915 nodule inhabiting bacteria were collected from white and subterranean clovers, grown in soils with different P fertilization histories. A subset (1,260) of these bacteria were used to determine that hydroxyapatite (HA) was most suitable substrate for in vitro screening. Of the 79 nodule bacteria that could solubilize P, 32% were rhizobia, 29% Enterobacter spp. and 17% Pseudomonas spp.. More PSB were found in subterranean clover (58) nodules than the white clover (21) nodules. Subterranean clover nodules also had more non-rhizobial PSB (44) compared with white clover nodules (10), suggesting that there may be an active recruitment by subterranean clover. PSB had a higher frequency in soils with high levels and duration of P fertilization, suggesting that the soil bacteria that could solubilize more P may have been selected. 16S rRNA gene amplicons sequencing (Illumina HiSeq) was used to characterise the bacterial communities of nodules, roots and rhizosphere of white and subterranean clovers, grown in different P fertilization histories. Bacterial communities were different in the two clover hosts and in the three P treated soils with 15 - 20% of nodule reads being non- rhizobial. More gammaproteobacteria, especially Pseudomonas and Enterobacter spp. were found in subterranean clover nodules. Nodule and root communities of subterranean clover were more affected by P fertilization than white clover. Root and nodule communities were more similar to each other than rhizosphere communities, suggesting that non-rhizobia entered nodules through the roots. P solubilization was quantified in liquid media and a decrease in pH was the main mechanism behind PS. Enterobacter and Pseudomonas had higher PS activity than Rhizobium spp.. The PS activity was unaffected by the presence of soluble P, suggesting that this activity may be a by-product of some other metabolic activity. Glucose was essential for the PS activity and only a few Enterobacter spp. could solubilize hydroxyapatite utilizing mannitol, indicating that these strains may have additional or alternate mechanisms for PS. Only 11 PSR could nodulate white clover, therefore, these strains were selected for further experiments. Of these, six increased (p

Perennial ryegrass, plantain, white clover and red clover were grown in monocultures and up to four- species mixtures in Canterbury, New Zealand. The dry matter production and weed content for the first year of growth (2015-2016) were measured. The species were sown based on a simplex mixture design that included all possible combinations of each species. Statistical analysis was performed on the data and a special cubic model was created. Mixtures were found to have greater annual yields and weed suppression than monocultures. All four of the monocultures had similar annual yields (average 16194 kg DM/ha), but they had different weed contents. The red clover and white clover monocultures had higher weed contents (average 15%) than plantain and ryegrass (average 3%). Success in the two-species mixtures was determined by the type of the ingredient species. Two- species mixtures containing a legume and a non-legume had greater annual yields (average 22849 kg DM/ha) than mixtures that contained two legumes or two non-legumes (average 17270 kg DM/ha). The white clover-red clover mixes and the ryegrass-plantain mixes performed similarly and were no better than the best performing monoculture. There was no difference in annual yield of ryegrass or plantain with white clover or red clover. All of the mixtures suppressed weeds well, except for the white clover and red clover two-species mixture, which was the only mixture to have a greater weed content than 5%, with a mean weed content of 18%. There was a yield advantage of including up to three species in a pasture mix, greater species richness than this prodivided no added yield advantages. There was greater weed suppression with two-species mixtures, but more species in the mix had similar weed contents. Evenness of species proportions in the mixtures had no effect on yield or weed supression. Ryegrass, plantain and white clover mixes were sown in alternate rows to test for an effect of species separation. There was found to be no difference in annual yield and weed content between mixtures that had species sown together and species sown apart. The model was used to get an optimum pasture mixture of 30% perennial ryegrass, 30% plantain and 40%, which gave a predicted annual yield of 25974 kg DM/ha.

An index to rank perennial ryegrass cultivars based on their relative economic benefit to pasture-based dairy systems, has been developed in New Zealand (Chapman et al., 2016; DairyNZ) in recent years. Performance values in this system are calculated for the key trait of seasonal dry matter (DM) production, using data from cultivar evaluation trials conducted using perennial ryegrass monocultures and high nitrogen (N) fertiliser inputs. To determine if the index should account for genotype × management interactions, experiments with a common design were established in four regions of New Zealand in 2012. Results of the first two years of the Canterbury experiment are presented in this thesis. The experiment was conducted at the Lincoln University Research Dairy Farm, Lincoln, Canterbury, and used a split plot design with eight subplots randomly allocated within four main plots each replicated in five blocks. Main plots comprised all combinations of pasture sown with or without white clover receiving either low (100 kg N/ha) or high (325 kg N/ha) rates of nitrogen fertilizer annually, randomized within blocks. In the plus clover treatments, pastures were sown with a 50:50 mixture of two clover cultivars commonly used in dairy pastures. The eight perennial ryegrass cultivars were selected to provide contrasting phenotypes for two traits that may influence competition between grass and clover: morphology, and heading date. The morphological contrast was between high tiller density/fine leaf material (both diploids) and low tiller density/broad leaf material (both tetraploids). The heading date contrast was between mid-season and late-season heading date materials, all of them diploids. Each main plot was grazed by dairy cows following standard farm management practices. Total DM yield was estimated in each subplot before grazing by cutting, using a Haldrup forage harvester. Botanical and pasture nutritive value sampling was conducted pre-grazing in spring, summer and autumn each year. Ryegrass and clover population density were measured in autumn each year. Results of the first two years of the experiment show that seasonal and total annual yield of the High N treatments was greater than from the Low N treatments. With the exception of spring of the establishment year, seasonal and total annual yield of the plus clover treatments was greater than from the minus clover treatments. N and clover interactions were observed in summer of the first year, autumn of both years, and for the total annual yield from both years. In general, the Low N plus clover treatment yielded similarly to the High N treatments, and yielded significantly more DM than the Low N minus clover treatment. The effect of cultivar on DM yield was significant in spring and autumn in both years, in winter 2013, and in the annual total of the second year. The clover content of pastures was always greater in the Low N treatments compared with the High N treatments and was affected by the ryegrass cultivar during spring in both years and in the second summer. There were no significant interactions between N and cultivar for clover content during the two years of the experiment. The heading date contrast affected the white clover content of pastures during summer in both years and in autumn 2013, resulting in mid heading date cultivars having greater white clover content than late heading date cultivars. Despite the effects of cultivar and treatments on DM yield and clover content, no significant interactions were detected between clover inclusion/exclusion and perennial ryegrass cultivar, or between N level and perennial ryegrass cultivar on seasonal or annual total DM yield, with the exception of winter 2013. As a consequence no evidence of re-ranking emerged and therefore performance values in the DairyNZ Forage Value Index (DairyNZ) do not need adjustment to account for grass-clover interactions over time and their effects on total pasture DM yield. The second experiment reported in this thesis was carried out with the objective of analysing how the perennial ryegrass and white clover characteristics affected their competitive ability, their proportion in the sward and the DM yield when grown in mixtures. The experiment was conducted at the Lincoln University Research Dairy Farm, Lincoln, Canterbury, and used a split plot design with four blocks. Main plots were two nitrogen levels (100 and 325 kg N/ha/year), randomised within blocks. Subplots were the pasture types (24), made up of a 4 × 4 factorial of 4 perennial ryegrass cultivars and 4 white clover cultivars (16 subplots) plus monocultures of each cultivar (8 subplots), randomised within main plots. Four perennial ryegrass cultivars were selected to create a range from fine to broader leaved materials and from open to denser cultivars. The four white clover cultivars were selected to create a range in leaf size, from small to large leaved. Total DM yield was estimated in each subplot by cutting, using a Haldrup forage harvester. Botanical composition was determined by dissecting a subsample collected from the harvested herbage at every harvest. Ryegrass and clover population density were measured four times during the experimental period (winter 2014 to autumn 2015). Photosynthetically active radiation (PAR) intercepted by the canopy and canopy height were measured three times during the year. Total annual DM yield of the mixtures was greater in the High than in the Low N treatment. The inclusion of clover increased the total annual DM yield under both N treatments, but the increment was greater under Low than under High N treatment. Only on one occasion was the white clover content of pasture affected by the interaction between perennial ryegrass and white clover cultivar, but not during the rest of the season. There were effects of perennial ryegrass and white clover cultivars on DM yield of the mixtures in some of the harvests during the year, but the total annual DM yield was similar for mixtures sown with different grass or different clover cultivars. With the exception of one occasion, no significant interactions were detected between perennial ryegrass cultivar and white clover cultivar on DM yield of the mixtures, meaning that the inclusion of different white clover phenotypes did not affect the DM yield of the mixture differently when associated to different perennial ryegrass phenotypes.

Nitrous oxide (N₂O) is a potent greenhouse gas (GHG) and the single-most ozone (O₃) depleting substance. Agriculture is the dominant source of anthropogenic N₂O emissions globally, and especially in New Zealand. Urine, synthetic nitrogen (N) fertiliser and farm dairy effluent (FDE) are the main sources of N₂O emissions from agricultural soils in New Zealand. Urine and synthetic N fertiliser have received considerable research attention to minimise their contribution to soil N₂O emissions due to the high N loadings and greater emission factors (EF) of these inputs. However, as the land application of Farm Dairy Effluent (FDE) is a less significant contributor to New Zealand’s overall N₂O emissions profile, research on this N-input is limited. The mass of FDE applied to land increased from, 18kt in 1990 to 39kt in 2013, and a recent increase in popularity of herd homes will further increase the mass of FDE produced. Thus, FDE requires further research. Limited data is available on the EF’s from the land application of FDE. In addition, an analysis of the literature suggests the NH₄⁺-N concentration of FDE is highly variable. The aim of this study was to determine the influence of FDE NH₄⁺-N concentration on soil N₂O emissions. A 35-day field trial was conducted, where 10 mm of FDE was applied to pasture at NH₄⁺-N concentrations of either 90, 150, 200, 300 or 400 mg NH4⁺-N L-1. The 150 and 400 mg NH₄⁺-N L-1 treatments contained 15N to monitor the fate of FDE NH₄⁺-N. N₂O gas samples were taken daily for the first week, then every 2-3 days for the remainder of the trial. Soil inorganic-N pools were monitored every 7 days. Pasture production was measured on days 19 and 35. Peak N₂O emissions occurred within 24 hrs of applying the FDE. The highest N₂O emisions were produced in the 400 mg NH4⁺-N L⁻¹ treatment averaging 65 kg N₂O ha⁻¹ day⁻¹. FDE treatments produced significantly more than the control until day 7. Emission factors ranged from 0.18 to 0.32 percent of total N applied, significantly less than the 1% currently used to calculate New Zealands GHG inventory. The emission period was relatively short due to low soil nitrate concentrations, and/or relatively dry soil conditions. Ammonium concentration is a key driver of N₂O emissions. Cumulative soil N₂O emissions increased linearly with FDE NH₄⁺-N concentration, however, it is likely this relationship may change in soils with a higher moisture content. Therefore, the use of a single relationship between the two variables for all environmental situations may not be possible. Further studies that analyse the influence of NH₄⁺-N concentrations on soil N₂O emissions, in a range of typical environments, are required to fully understand this relationship.

In organic soybean– winter wheat – corn rotations, animal manure is a common choice to maintain high yields, but leguminous crops grown as green manures after wheat harvest and incorporated into the soil before corn planting, can be an alternative when animal manure is not accessible. Forage legumes with high dry matter (DM) production and high biological N fixation have been shown to meet corn N demand. However, in Eastern Nebraska, lack of precipitation can reduce green manure growth and N fixation, leading to an insufficient N supply for corn, but corn growth can also be impacted by green manure soil water use. Our objectives were 1) to determine the green manure potential of four forage legumes, and 2) to evaluate management methods that optimize green manure benefits. We conducted an experiment at the ARDC near Mead, NE, from 2011 - 2014. Red clover, white clover, alfalfa, and sweet clover were undersown into winter wheat in early spring. After wheat harvest, they were either mowed or not mowed, and terminated in the fall or the next spring. We measured green manure DM, weed DM, soil nitrate concentrations, and crop yields throughout the rotation. We compared green manure effects to effects of cattle manure, post-wheat soybean green manure, and a control (no fertilizer). Red clover produced the most DM, up to 5.5 Mg ha-1 and showed excellent weed control, especially when mowed. Green manures did not increase soil N compared to the control. Corn yields were always significantly higher after cattle manure (7.6 to 8.1 Mg ha-1) than after undersown green manures, and were lowest after red clover in 2012 (2.8 Mg ha-1) and after white clover in 2013 (4.6 Mg ha-1), because of the clovers’ high soil water use and insufficient N production. In our study, green manures established well, but increased corn yields compared to a control in only one of three years. Cattle manure was the most reliable method to maintain high crop yields. Future research should investigate combinations of cattle and green manure to increase N availability to corn and decrease N leaching losses after corn harvest. Advisor: James R. Brandle

Sediments from the Chatham Rise, a by-product from phosphate nodule extraction, could have some agronomic value for New Zealand pastoral soils. Two pot trials were carried out to investigate the agronomic effectiveness of four selected sediments with varying levels of nutrients, mainly calcium (Ca) and phosphorus (P). The first experiment tested all sediments at a range of application rates in contrasting soils growing Italian ryegrass (Lolium multiflorum). The second experiment tested one selected sediment at a constant rate of P against Chatham Rise phosphate rock (PR) and triple superphosphate (TSP), growing monocultures of Italian ryegrass and white clover (Trifolium repens). Results showed that sediment application did not increase ryegrass dry matter yield or P uptake, however did have a liming effect. One sediment, DD44W, actually reduced clover dry matter yield in the second experiment, and P uptake was significantly reduced compared to PR and TSP. The findings demonstrate that these sediments are of limited value as a P fertiliser, however the sediments with higher Ca content can be a useful liming material. The implications are that if the mining project gets consent, these sediments could be retained rather than returned to the sea floor after phosphate nodule extraction. These findings could also be beneficial in the consenting process and could provide a useful product for New Zealand farmers while also providing a solution that reduces the environmental impacts of the sea floor mining operation.

The majority of New Zealand pastoral agriculture is based around white clover and perennial ryegrass. These pastures a generally well suited to the majority of New Zealand. However these pastures often fail to perform at crucial periods, such as in summer when pasture growth and quality is vital to finish livestock. The inclusion of summer active pasture species with high pasture quality could provide a solution to poor summer production from white clover-perennial ryegrass pastures. A field experiment was conducted to investigate the effect of different pasture mixtures and sowing strategies on the yield and composition of sown species over an establishment period of 7 months from an autumn sowing. The sown species included tetraploid ryegrass, white clover, plantain and red clover were used in a number of pasture mixture combinations. The experiment was conducted at Lincoln University, Canterbury, New Zealand from the 26th of March to the 22nd of September 2015. Accumulated yield from sowing to the 22nd of September was significantly different over the 23 pasture mixtures. Sowing pasture mixtures with proportions, of tetraploid perennial ryegrass and/or plantain of 50% or more results in high accumulated yields between 3000 – 3900 kg DM/ha. White and red clover pure swards produced poor yields

Aluminium (Al) toxicity is a key factor limiting pasture production in the high country of New Zealand. Many high country soils are acidic, with a pH less than 5.5. Low soil pH makes Al soluble, leading to high concentations of exchangeable Al in soils and plant Al toxicity. Aluminium toxicity reduces plant vigour and yields; resticting root growth and limiting nutrient and water uptake. The harsh climatic conditions experienced in the high country could also play a role in the exacerbation of Al toxicity. Literature on the influence of low moisture extremes on exchangeable Al concentrations is scarce. Nomad white clover was grown for 10 months under glasshouse conditions at Lincoln University in two acidic high country soils collected from Omarama and Glenmore Stations, Central Otago. The trial was a full factorial design of two soil types, three soil moisture levels and four fertiliser treatments. Lime (100% CaCO3) was applied at three rates (0,2 and 8 t lime/ha) with one treatment of P and lime (2 t lime/ha and 150 mg P/kg). Sulphur was applied to all treatments at 120 kg S/ha (Gypsum). Plants were harvested four times during the experiment and yields were determined. Herbage was analysed for macro and micro nutrient concentrations by ICP-OES and NIRS. Soils were analysed for pH and exchangeable Al, with bulk soil samples also analysed for P and S contents at completion of the experiment. A significant interaction between soil moisture and exchangeable soil aluminium was observed for the Glenmore soil. The medium moisture treatment (23% VWC) had lower exchangeable Al, 4 mg/kg compared to the low (17% VWC) and high (33% VWC) moisture treatments which contained 5 and 5.6 mg exchangeable Al/kg respectively, for the zero lime treatment. The low Al Omarama soil in comparison, only contained Al concentrations of 1.3, 1.1 and 0.8 mg Al/kg in the same treatments at the completion of the experiment. This experiment did show an interesting interaction and strongly suggests that soil moisture influences exchangeable Al, but only when Al concentrations are greater than 4 mg Al/kg. Similar experiments are required using high Al soils to confirm this result. The Glenmore soil contained higher initial concentrations of exchangeable Al (6.6 mg Al/kg), had greater carbon stocks (6.9 kg C/m3) and produced higher shoot yields (10.6 g DM/pot) than the Omarama soil (3.8 mg Al/kg , 3.5 kg C/m3 and 7.8 g DM/pot). The high P fertility of the Glenmore soil (Olsen P 18) probably supressed soil Al affects on yield. High soil carbon may have also masked the effects of soil moisture on exchangeable Al for the Glenmore soil. Moisture treatments caused a significant growth limitation on the low moisture treatment (17% VWC), with a yield of 8.7 g DM/pot compared to 9.5 and 9.4 g DM/pot for the medium (23% VWC) and high (33% VWC) moisture treatments. Herbage nutrient concentrations were effected predicatably by lime and P addition. Shoot concentrations of P and Mo increased with lime addition, while shoot concentrations of Zn, Mn, B and Al decreased. Shoot nutrient uptake followed the same trends as concentration.

Bermudagrass (Cynodon dactylon L.) is a perennial warm-season grass that is very responsive to nitrogen (N) fertilization. Excessive N applications have negative environmental consequences and make maintaining mixed swards difficult. This study determined the effects of enhanced efficiency (EE) N fertilizers and fertilizer rate on bermudagrass yields, nutritive values, and white clover persistence. Nitrogen sources included urea, urea formulated with Agrotain® (U+A), urea with Agrotain® and dicyandiamide (SuperU), a polymer-coated urea (ESN), ESN+urea (75% ESN, 25% urea), and methylene urea (MU). In the urea formulation trial, SuperU and U+A maximized forage yields at lower N rates. The dicyandiamide in SuperU did not increase yields over U+A. Highest production efficiency was achieved at lower N rates. ESN had the lowest relative stimulate growth, which may increase clover persistence. In the clover persistence trial, the addition of N fertilizer began decreasing clover populations after 112 kg N/ha. Slow-release fertilizers (ESN, ESN+urea, MU) had higher clover percentage at the final harvest. Crude protein and in vitro digestible dry matter increased, while neutral detergent fiber and acid detergent fiber decreased with the addition of white clover. SuperU and U+A were more efficient and ESN had lower relative stimulated growth beneficial for mixed pastures.

This study involved three key experiments to examined the effect of Nil endophtye and novel endophyte strains (AR1, AR37, Wild-Type (WT) and NEA2) of perennial ryegrass on dry-matter (DM) production, botanical compositon, alkaloid profiles and possible effects on soil-pasture ecosystems interactions. A field trial measured DM production and composition of perennial ryegrass pastures sown with a diversity treatment of with or without additional prairie grass, herb (chicory) and legume (white clover). In the same trial, alkaloid and endophyte concentrations were examined in dissected vegetative grass tillers to ascertain how alkaloids may potentially leave the plant and affect soil-pasture ecosystems. Finally, a pot trial comparing the addition of live or dead perennial ryegrass plant material was used to examine the effect of endophyte strain on the growth of associated pasture species. There was no effect of endophyte or diversity treatment on the total or seasonal herbage DM production of the field trial, with a non-significant 18,540 kg DM/ha and 18,090 kg DM/ha total yield for the diverse and monoculture treatments respectively over the 16 months of the trial. There was a significant endophyte x diversity interaction in spring 2011 when monocultures of Nil and AR37 (5638 and 4963 kg DM/ha respectively) yielded significantly higher than that of their equivalent diverse pastures (4099 and 4323 kg DM/ha). It is unclear the reasoning for this interaction. The different seasons showed significant differences in the proportion of white clover in the diverse swards with the highest in summer, and the lowest in winter (25% and 11% respectively as a porportion of the total sward DM). There were minimal effects of endophyte on botanical composition, with no significant differences for monocultures. The proportion of prairie grass present in summer 2010/11 was significantly higher in WT diverse plots than other endophytes and spring 2011 showed higher chicory present in WT and NEA2 diverse plots over AR37. Endophyte concentration differed for the different disected parts of the grass tillers and different endophyte strains. The highest concentrations of endophyte was in the stem with 317 copies per ng of gDNA. The leaves ranged from the youngest first leaf with 59 copies, increasing with leaf age to 268 copies in the remnant leaf. This is consistent with other findings of the endophyte being located mainly at the base of the plant and having hyphae carried up as the leaf expands. NEA2 had lower endophyte concentration with roughly 50% less copies of endophyte gDNA/ng than the next closest treatment (AR1), approximately 66% less than WT, and 68% less than AR37. NEA2 had significantly less peramine than WT or AR1 for all tiller parts. The pot trial live and dead plant material thatches differed in alkaloids. Peramine and Janthitrems were higher in the live than the dead plant material. Lolitrem B concentrations for NEA2 were similar in the live and dead plant material (1.55 and 1.34 ppm respectively), but higher in WT endophyte in the dead compared to the live plant material (4.43 vs 1.04 ppm). Ergovaline was low (

A series of experiments were conducted to estimate the genetic parameters for an elite breeding population of white clover (Trifolium repens L.) in both an irrigated and dryland environment. Data collected from two environments using spaced-planted and mini-plot trials were used to demonstrate the rationale for pursing various traditional and modern marker-assisted selection breeding methods in white clover. In addition, and under the over-arching theme of investigating breeding strategies, supplementary findings are presented regarding pollination patterns within a white clover population in isolation cages. In an experiment designed to generate half-sib and full-sib families for progeny evaluation, pollen dispersal within isolation cages using wild bumble bees (Bombus sp.) was monitored using simple sequence repeat molecular markers (SSR markers). Within bi-parental crosses, no detectable levels of self-fertilised germinated seedlings were detected, nor were there any detectable levels of foreign pollen sources. Within polycross isolations, no detectable levels of foreign pollen sources were detected, and self-fertilised rates were negligible (

Pasture production and composition and liveweight gain of sheep grazing tall fescue-clover and perennial ryegrass-clover dryland pastures at Lincoln, Canterbury was examined in four experiments from May 2008 to December 2010. Four objectives were identified to (i) quantify the pasture production, growth, persistence, composition, (ii) pasture nutritive value and liveweight gain of sheep grazed cultivars of tall fescue, white, strawberry and subterranean clover at high and low P and S fertility, to determine the (iii) effect of time of grazing in autumn on subterranean clover growth was determined, and (iv) effect of spring defoliation management on tall fescue cultivars and perennial ryegrass pastures grown with white and strawberry clover. Experiment 1 examined the interactive effects of tall fescue cultivar (Continental, cultivar Advance, versus Mediterranean, cultivar Flecha), sown clover species (subterranean clover versus a combination of white and strawberry clover) and soil fertility (high P and S fertilizer versus low fertilizer P and S) on annual and seasonal DM production, botanical composition, plant density and morphology. Annual DM production was greater in Advance (7790 kg DM/ha) than Flecha (6871 kg DM/ha) in year 1 and Flecha (8367 kg DM/ha) than CA (7547 kg DM/ha) in year 2. Annual DM production of the legume component was greater in subterranean clover pastures than perennial clover pastures in both year 1 (945kg DM/ha v 539 kg DM/ha) and year 2 (1457kg DM/ha v 1042 kg DM/ha). Survival of strawberry clover in perennial clover pastures was poor with a low percentage (< 1 % DM) in pasture after 2 years. The percentage of clover in pastures was generally greater in Flecha than Advance. Annual DM production was greater at high than low fertility in both year 1 (8734kg DM/ha v 7828 kg DM/ha) and year 2 (7134 kg DM/ha v 6834 kg DM/ha). This primarily reflected greater grass growth at high fertility as legume growth was unaffected by soil fertility. Experiment 2 examined the interactive effects of tall fescue cultivar (Advance versus Flecha) and soil fertility (high P and S fertilizer versus low fertilizer P and S) on pasture nutritive value and liveweight gain of ewes and lambs in spring and ewe hoggets in autumn. Liveweight gain per head of single lambs grazing with lactating ewes (22 ewes/ha) was 284 g/head/day over a 27 day period in spring of year 1. Liveweight gain per head of twin lambs grazing with lactating ewes (25 ewes/ha) was 258 g/head/day over a 60 day period in spring of year 2. Over the total spring period in each year, there was no effect of soil fertility or tall fescue cultivar on liveweight per head or ha of lambs. This occurred despite pasture growth and feed availability being on occasions greater in Flecha than Advance, and at high than low soil fertility. Liveweight gain per head of ewe hoggets (22 ewes/ha) was greater in Flecha (52 g/head/day) than Advance (14 g/head/day) over a 51 day period in autumn of year 1 but greater in Advance (282 g/head/day) than Flecha (197 g/head/day) over a period of 18 days in autumn of year 2. The different cultivar effects between years reflected the availability of feed in response to summer rainfall at the start of the grazing period as well as pasture growth during the grazing period. The clover percentage was average 20% and 31.2% in spring 2008 and 2009, respectively. Experiment 3 examined the effects of grazing subterranean clover at the spade leaf, three trifoliate and six trifoliate leaf stages in autumn on seedling populations, seedling survival and DM production of subterranean clover when grown with Advance and Flecha tall fescue cultivars. Seedling death during grazing was greater at the spade leaf (43%) than the three leaf or six leaf (14%) stage. However, because seedling populations were initially high (1326 plants m2), this high death with early grazing simply killed seedlings that would have died for other reasons; and seedling numbers at the start of winter were similar among treatments (619 plants m2). Spring clover DM production was greater at the spade leaf (2834 kg DM/ha) than three leaf (2573 kg DM/ha) or six leaf stage (2648 kg DM/ha), reflecting greater plant size than any difference in plant population. Delaying grazing until after the three leaf stage will enhance seedling survival when seedling populations are low. Experiment 4 examined DM production, botanical composition, and plant morphology of perennial ryegrass, Advance and Flecha tall fescue pastures grown with strawberry clover and white clover under continuous stocking and rotational grazing in spring. DM production in the establishment phase was greater in perennial ryegrass (1098 kg DM/ha) than Advance (563 kg DM/ha) or Flecha (503 kg DM/ha). In the first full year, DM production was greater in perennial ryegrass (9417 kg DM/ha) and Advance (9928 kg DM/ha) than Flecha (7559 kg DM/ha). Annual DM production of legume was greater in white clover (1638 kg DM/ha) than strawberry clover (670 kg DM/ha). Persistence of strawberry clover was poor with the plant population declining from 145 plants/m² at start to 9 plants/m² after 20 months. Although taproots were larger and lasted longer in strawberry clover than white clover plants, strawberry clover plants had fewer nodal roots, which appeared to limit persistence under grazing. There was no effect of continuous stocking versus rotational grazing on DM production or persistence of white and strawberry clover. The study showed that pastures based on either Continental or Mediterranean tall fescue cultivars are capable of high DM production in dryland pastures in Canterbury, particularly when grown with subterranean clover. These pastures gave high liveweight gain, which would enable lambs to be finished in spring before the onset of summer dry conditions. Under the conditions of this study, there was little detectable effect of tall fescue cultivar on livestock performance, indicating a role of both types of tall fescue for dryland pastures in Canterbury. Strawberry clover persisted poorly under all grazing management, and without breeding for improved nodal root development or revised grazing management guidelines, will not be a suitable legume for dryland pastures.

Increasing legume abundance is an important component of pastoral intensification, in providing increased quality feed and nitrogen inputs to nitrogen deficient, hill and high country grassland. Establishment and persistence of traditionally sown clover species white clover (Trifolium repens L.) and subterranean clover (Trifolium subterraneum L.) is often limited in summer dry areas. The ecology of naturalised adventive (volunteer) annual clovers species Trifolium arvense L., Trifolium dubium Sibth, Trifolium glomeratum L., and Trifolium striatum L. commonly found in the South Island hill and high country areas of New Zealand was examined through a series of site surveys, field trials, and glasshouse experiments. The relationship of these clover species to environmental factors of climate (annual rainfall), topography (slope, altitude, aspect), soil fertility (phosphorus, sulphur, pH), and sheep grazing (spring management, and preference) were investigated. On two contrasting hill and high country stations in the South Island: Glenfalloch in inland Canterbury (1665 mm annual rainfall) and Mt Grand in Central Otago (703 mm annual rainfall), the abundance of these species was quantified in relation to environmental factors. Site surveys were conducted in three hill blocks per farm, with measurements of grassland species cover, slope, aspect, grazing intensity, soil fertility, soil depth, and altitude, made within quadrats along three transects at upper, middle and lower hill slope positions. The single adventive clover present at Glenfalloch was Trifolium dubium, whereas all adventive species were present at Mt Grand. The % cover of adventive annual clovers was greater (30.1) than that of white clover Trifolium repens (3.8) or subterranean clover Trifolium subterraneum (0.1) at Mt Grand. Trifolium repens % cover decreased with increasing altitude. Striated clover Trifolium striatum and cluster clover Trifolium glomeratum % cover was greatest on sunny aspects, while Trifolium repens % cover was greatest on shady aspects. Trifolium repens cover increased and Trifolium striatum cover decreased with increasing available soil phosphorus (P). Annual rainfall appeared to be the main environmental factor influencing the abundance of adventive annual clover species on two geographically and climatically different South Island high country stations. A two year spring grazing management and superphosphate (SP) application field trial was conducted at a typical summer dry pasture site on a mid altitude (~700 m a.s.l.) north-facing moderately steep, hill site at Mt Grand Station. Adventive annual clover biomass, botanical composition and population dynamics were measured from October 2008–May 2010, from low (75 kg/ha) and high (200 kg/ha) SP application split plots of 2.5 x 5 m area within different grazing management main plots (continuous and mid-spring closure) of 5 x 5 m area. Adventive annual clover species contributed the largest proportion of legume sward content at this site, with sown species contributing the least. Collectively, adventive annual clover species accounted for over 90% of the sward clover content present during both years. Trifolium striatum was the dominant adventive species. Large variation in biomass production (668 and 51 kg DM/ha), sward botanical contribution (28 and 2%) and seedling recruitment (9.7 and 24 seedlings/core) by adventive annual clovers was observed between the two years of sampling, which was attributed to high rainfall during spring and early summer of the first year compared to the very dry spring of the second year. Deferred spring grazing resulted in significantly more adventive annual clover biomass in a moist year (1452 kg DM/ha) in comparison to continuous grazing (674–805 kg DM/ha). However, this did not appear to result in significantly more adventive annual clover seed production, seedling recruitment, or biomass in the following year. Superphosphate (SP) application had no significant overall influence on adventive annual clover species biomass. High SP application did have a negative effect on autumn seedling recruitment of the three most prevalent adventive annual clover species (Trifolium striatum, T. dubium, and T. glomeratum). Soil plant-available N sourced from pasture legumes has been shown to be positively associated with long-term use of fertiliser phosphorus (P) and sulphur (S) use in NZ high and hill country. Determining the nutrient requirements of little known naturalised, adventive pasture legume species such as cluster clover (Trifolium glomeratum), haresfoot clover (T. arvense), striated clover (T. striatum) and suckling clover (T. dubium) that exist on low fertility, summer dry hill and high country slopes is important for the productive pastoral sustainability of extensive livestock grazing agroecosystems. The first glasshouse study was conducted to determine the yield response of these four adventive annual species, plus Trifolium subterraneum and T. repens, to increasing levels of available P in a typical, low fertility, acidic NZ high country soil. The order of greatest yield DM response was T. subterraneum > T. arvense > T. repens > T. dubium > T. glomeratum > T. striatum (4.4–0.8 g DM/pot), while the P application rates at which maximum yield was produced varied between the species with T. arvense and T. striatum yielding the most DM at 250 mg P/kg soil, T. subterraneum, T. glomeratum and T. repens yielded the greatest amount at 500 mg P/kg soil, and T. dubium producing its highest yield at 2500 mg P/kg soil. The order of greatest P-response efficiency by species (how much P was required for each gram of DM produced) was T. subterraneum > T. arvense > T. repens > T. glomeratum > T. dubium > T. striatum. Implications for low input, extensive grazing systems in hill and high country areas are discussed. The growth response and nutrient uptake of four adventive annual clovers to applied S or lime (CaCO3), grown in a typical low fertility South Island (S.I.) high country soil, were investigated under glasshouse conditions and compared to white and subterranean clovers as reference species. The annual species had yield responses of 12–17%, or were unresponsive, to S applications. Maximum yields were generally in the order of T. subterraneum ≥ T. arvense > T. striatum ≥ T. dubium > T. glomeratum > T. repens (8.0–5.2 g DM/pot). For lime treatments, yields were strongly driven by P availability, linked to soil pH. Trifolium repens and T. striatum responded to liming at low lime rates (1 t/ha equivalent), while all other species had negative yield responses to liming. The data indicate that the adventive annual clovers are better adapted to low soil fertility (low pH and S) conditions, which in turn may be an important factor contributing to their success under S.I. high country field conditions. Grazing preference of Merino sheep for the adventive annual clover species was investigated in a field trial at a dryland pasture site on the Canterbury Plains at Lincoln University. The quantity of clover species sward biomass removed during grazing and the time spent grazing from different clover species swards was quantified during grazing preferences tests in November and December. Relative preference (Chesson-Manly index) for each clover species sward on offer was calculated from the utilisation of each species relative to the utilisation of all other species on offer. Relative preference decreased and was significantly different between species when they reached a reproductive stage of maturity. This was attributed to an overall decrease in nutritive value of clover herbage, with relative preference significantly correlated with the concentration of acid detergent fibre (ADF), neutral detergent fibre (NDF), protein, and dry matter digestibility (DMD) chemical components. The latest flowering clover species were the most preferred, with Trifolium repens the most preferred clover species overall. Trifolium arvense was the most preferred adventive clover species when most species were predominately vegetative. Implications of key results and further research avenues are discussed.

A series of experiments were conducted to determine the effect of hybridisation with Trifolium uniflorum L. on root and shoot characteristics of Trifolium repens L. (white clover). In a grazing experiment, dry matter (DM) yield scores combined over the experimental period were higher for white clover (5.3) than for BC₁ (backcross 1) (4.4) and BC₂ (backcross 2) (4.3) hybrid generations. The proportion of nitrogen from fixation (85–96%) was not affected by hybridisation, and there were generally no differences among clover types in shoot %N. In 13 month old plants, tap root survival was higher for T. uniflorum (50%), and BC₁ (31%) than for BC₂ (13%) and white clover (11%). Some tap roots of T. uniflorum and the BC₁ generation survived up to 19–20 months, but those of BC₂ and white clover did not. In contrast to the field experiment, shoot dry weight (DW) of BC₁ hybrids was 2–24 times higher than white clover over four harvest times in a glasshouse root tube experiment using sand culture with a low ionic strength (LIS) nutrient solution. Shoot P concentrations (925–1716 mg kg⁻¹) were below critical levels for white clover growth. In a pot experiment with the same LIS treatment, these effects were confirmed. Over a range of nutrient treatments, some hybrid families were less affected by decreases in nutrient solution strength than others. In all treatments combined, shoot DW of Kopu II x 900-4 was 2.8 times higher than that of the Kopu II parent, and Kopu II x 487-9 was 1.5 times higher. Crusader hybrids did not differ to the Crusader parent. In some treatments, P- and Pi-use efficiency of Kopu II x 900-4 was higher than the other clover types. Some root characteristics of T. uniflorum, inherited by BC₁ hybrids, may affect water and nutrient interception. In a glasshouse tube experiment, BC₁ hybrids and T. uniflorum had more root mass in the upper part of the profile than white clover. In a hydroponics experiment, the roots of T. uniflorum were thicker than BC₁ hybrids and white clover, which may contribute to drought resistance. Topological indices of T. uniflorum (0.91–0.94) were higher than for white clover and some hybrid families (0.76–0.92), as were those of Kopu II x 900-4 (0.85–0.98). This herringbone root architecture may be adaptive to low soil fertility. In a rain shelter experiment, shoot DWs of white clover and BC₂ were 95% and 26% higher, respectively, than that of BC₁ in the Watered treatment. However, shoot DW decreased less under water stress for the BC₁ generation (47%) compared with BC₂ and white clover (nearly 70%). There were no differences in shoot DW among clover types in the Stressed treatment. Stolon morphological characteristics of BC₁ (-31–68%) also decreased less under water stress than those of BC₂ (-44–73%) and white clover (-38–74%). Kopu II BC₁ was able to maintain photosynthesis and transpiration under lower leaf water potentials than Kopu II BC₂ and Kopu II. Net photosynthesis of Kopu II BC₁ did not change under water stress, but in Kopu II BC₂and Kopu II it decreased by 48% and 44%, respectively. There was also no change in transpiration with water stress for Kopu II BC₁, but in Kopu II BC₂ and Kopu II it decreased by 60%, resulting in a higher mean transpiration rate for Kopu II BC₁ in the Stressed treatment. Under water stress, the leaf water potential of Kopu BC₁ decreased more (-47%) than that of Kopu II BC₂ (-28%) and Kopu II (-31%). Root DW of Kopu II BC₁ increased by 59% under water stress, but that of Kopu II BC₂ and white clover did not change significantly. In the Stressed treatment, root DW of Kopu II BC₁ was 72% higher than that of Kopu II BC₂, and mean cross-sectional area of the thickest nodal root of Kopu II BC₁ was 2.6 times higher than Kopu II. In addition to improvements for some traits in broad hybrid generations, there was also evidence of differences among hybrid families for many parameters, which should enable selection of superior families and identification of segregating populations.

Perennial ryegrass pastures (Lolium perenne L.) were sown on 27 March 2012 with eight different cultivar treatments of a perennial ryegrass varying in ploidy (diploid or tetraploid), maturity (early or late flowering) and phenotype (open, broad leaf or fine, dense leaf) characteristics. A 50:50 mix of two white clover (Trifolium repens L.) cultivars was hand sown on 28 March in all plots 2012. Ryegrass and white clover herbage production, botanical composition, seedling development, shoot: root ratios, plant numbers and clover morphology of autumn sown pastures were assessed with particular reference to maturity and phenotype appearances. There was no advantage in terms of dry matter production from sowing different cultivar treatments. In September the dry matter production of the treatments ranged from 1529.77 kg DM/ha (One 50) to 1882.76 kg DM/ha (Commando). However, these results were not significantly different. In September, Abermagic treatment contained the highest amount of weeds, however it is unknown why. Kamo and Commando treatments contained a much higher percentage of dead matter in the sward compared to One 50 which contained the least. Perennial ryegrass content and white clover content in the swards did not differ significantly in any of the treatments. Significant differences were also found between development and cultivar however, it was unclear why this result occurred as it contradicted what was expected. Overall, cultivar effects on white clover do not occur in the first year of establishment due to ryegrass cultivars not mature enough to express characteristics that may influence the growth of white clover in the sward. Sowing rate, nitrogen application and severity of grazing should be incorporated into future studies to understand more about clover content in different ryegrass cultivar swards. Future studies after establishment should give more significant differences between ryegrass cultivars due to differences in characteristics such as leaf size, heading date and ploidy.

Introduction. White clover (Trifolium repens L.) is an out-crossing allotetraploid (2n=4x=32) that acts like a diploid, also called amphidiploid. White clover is a forage crop with a high level of genetic heterogeneity, which allows it to adapt to a wide range of environments. However, growth of white clover is often impaired by summer drought. New cultivars need to adapt to higher levels of abiotic stress to survive and persist. Recent research in white clover populations demonstrated that levels of a secondary metabolite, glycosides of the phenolic compound quercetin (Q), were positively associated with abiotic stress resistance, and inversely related to dry matter (DM) production. Molecular plant breeding methods using DNA markers and quantitative trait loci (QTL) analysis can increase breeding efficiency using information on genotype to complement phenotype, and identify genes controlling complex traits. Aims. The objectives of this study were (i) to investigate genetic control of the complex traits Q glycosides and DM production in contrasting environments, (ii) to test the inverse correlation between Q glycosides and DM production, and (iii) to identify DNA markers associated with quantitative trait loci (QTL) for use in marker-assisted selection (MAS). Materials and methods. An F1 mapping population (n = 190) created from a paired cross between two highly heterozygous white clover genotypes sampled from New Zealand cultivar “Kopu II” and Chinese ecotype “Tienshan” was used for phenotyping and genotyping. Phenotyping consisted of three parts: 1) in pots under outside conditions in separate pilot and phenotyping experiments, 2) in pots outside under imposed water deficit, and 3) in the field in two environments contrasting in soil moisture. Important traits were measured, including morphological, e.g. root DM, shoot DM, total DM, root-to-shoot DM ratio, stolon density, leaf size; biochemical, glycoside levels of the flavonols quercetin and kaempferol (K), Q:K glycoside ratio (QKR); as well as physiological, e.g. water potential, stomatal conductance and carbon isotope discrimination. Genotyping was investigated by means of single sequence repeat (SSR) markers and diversity array technology (DArT) markers from the genomes of Trifolium repens, Trifolium occidentale and Medicago truncatula. Results. Phenotyping in pot studies revealed sufficient genetic variation in the population, and only a weak albeit significant (P< 0.001) correlation between dry matter yield and Q, with levels of one explaining less than 10% of variation in the other. Imposed water deficit decreased leaf water potential by more than half overall, Q glycosides increased more than twofold and the ratio of quercetin to kaempferol glycosides increased preferentially towards Q. Furthermore, the most productive genotypes in the controls showed the greatest proportional reduction, the root:shoot ratio increased by half and individuals with high QKR levels reduced their biomass least under water deficit, and in turn increased their Q glycosides and QKR most. In the field studies, water deficit significantly reduced carbon isotope discrimination, Q glycosides were preferentially synthesised over K glycosides, accumulation of Q glycosides was related to retaining biomass, and stolon density was inversely related to stolon numbers. Genotyping resulted in a large number of DNA markers associated with genome regions influencing traits in the population. QTL effects were consistent over pot- and field experiments in some cases. Suitable markers were used to build a partial linkage map to better test for chromosomal locations containing QTLs associated with the traits of interest. Interaction between Q glycosides and DM QTLs was analysed. Discussion and conclusions. The data show, for the first time, that at the individual genotype level increased Q glycoside accumulation in response to water deficit is associated with retaining higher levels of biomass production. These findings are the first indication that forage populations that are both high yielding and show high Q glycoside levels are possible. The single locus discoveries of marker-trait associations provide a basis for enhanced plant breeding efficiency for these traits, and specific sources of new variation for economically significant traits. Overall, the findings can be used to better understand the physiological underpinnings of water stress relations in forage plants, and improve gain from selection in white clover by increasing the precision with which improved plants and populations are identified.

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.

Despite the importance of a high white clover (Trifolium repens) content in temperate pastoral systems in terms of livestock performance and nitrogen fixation, the proportion of white clover in grass-clover pastures is often low (

A trial was conducted on an irrigated site at Lincoln University, Canterbury to analyse the effects of new novel endophyte strains in perennial ryegrass cultivars and nitrogen on pasture production and composition. Novel endophyte strains were AR1 and AR37 (control: nil endophyte) with ryegrass cultivars Samson (diploid) and Aberdart·(diploid, expressing a high water soluble carbohydrate trait). Combinations of novel endophyte and ryegrass cultivar were tested (ALE, AAR1, SLE, SAR1, SAR37). Two nitrogen fertiliser regimes were applied to plots, low N (0 kg N/ha/year) and high N (160 kg N/ha/year). Total dry matter production during the trial period (February to September 2008) was highest in SLE with 3661 kg DM/ha and lowest in ALE with 3249 kg DM/ha. No significant differences in total or seasonal pasture production were found between endophyte x ryegrass cultivar treatments (p>0.05). There was a trend to low clover content in SAR37 (mean, 7.0%), however this was not significantly lower than AR1 or low endophyte treatments (p>0.05). SAR37 and AAR1 produced the lowest clover plant densities (172 and 183 plants/m² respectively; p=0.023) and clover growing point densities (612 and 798 GP/m² respectively; p=0.017) in autumn. Tiller density was higher in endophyte infected treatments in autumn with 10840, 7064 and 8157 tillers/m² in AAR1, SAR1 and SAR37 respectively. A negative relationship between tiller density and white clover plant density was found in autumn (R²=0.47), and produced an apparent tiller density threshold of around 6000 tillers/m² below which clover density was reduced to below 400 plants/m². Nitrogen fertiliser (160 kg N/ha/year) reduced clover content from 23 (no-N) to 7% in April, 7 to 2% in June and 23 to 7% in September. No evidence of interactions between endophyte, cultivar and nitrogen were found in any season (p>0.05). There was no evidence of difference in clover content due to more selection by sheep in some endophyte plots than others, thus diet selection was eliminated as a determining factor of low clover content. Instead, higher tiller density in endophyte infected pastures appears to suppress clover plant density. Therefore, an increased competitive advantage of endophyte infected ryegrass is suggested as the main factor causing reduced clover content in pastures, however, there is more research required in this area before conclusions can be made.

This research compared pasture establishment, with or without cultivation, with two drill types, Cross Slot or triple disc and with 0, 2 or 6 t ha ⁻¹ of lime in an herbicided, undeveloped depleted tussock grassland, at Lees Valley, Canterbury in autumn 2007. The objective was to identify the most cost-effective method of establishment. On the 2nd October, the cultivated treatment had total green dry matter production of 656 kg DM ha⁻¹, which was 2.25 times greater than the non-cultivated treatment of 292 kg DM ha⁻¹. Cultivation physically removed competing resident vegetation and resulted in greater mineralisation of nitrogen. The Cross Slot drill had total green dry matter of 542 kg DM ha⁻¹, which was 1.34 times greater than the triple disc drill of 406 kg DM ha⁻¹. The Cross Slot drill resulted in improved uptake efficiency of nutrients due to fertiliser placement. Total green dry matter production was 555 kg DM ha⁻¹ for 6 t of lime ha ⁻¹ and ~430 kg DM ha⁻¹ for 0 and 2 t of lime ha⁻¹. The application of 6 t lime ha⁻¹ increased soil pH by ~ 0.4 units. Pasture response was due to alleviation of A1 toxicity, increased plant availability of N and possibly Mo or a combination of these. Drill type and the rate of lime had no effect on plant population. By the 29th August ryegrass plant population in the cultivated treatment was 346 plants per m² which was 1.25 times greater than the non-cultivated treatment of 278 plants per m². Subterranean clover plant population was 66 plants per m². Ryegrass made up 40-67% of total green dry matter, subterranean clover 21-33%, white clover 3-6% and cocksfoot > 1.5%. The cultivated, the Cross Slot drill and the application of 6 t lime ha⁻¹ treatments compared with the non-cultivated, triple disc drill and 0 t lime ha⁻¹ treatments respectively reduced clover content due to an increase in ryegrass yield rather than a decrease in clover yield. Cultivation treatment and drill type affected the rate of ryegrass tiller appearance and therefore, tiller number per plant and per unit area. For the cultivated treatment, ryegrass tiller appearance between 500 and 1430 °C d averaged every 39 °C d compared with every 76 °C d for the non-cultivated treatment. For the Cross Slot treatment, ryegrass tiller appearance averaged every 44 °C d compared with every 63 °C d for the triple disc treatment. Increased tiller number plant⁻¹ possibly increased radiation interception, the driver of crop yield. The cultivation treatment had the greatest cost benefit due to the highest increase in production. The difference between cultivation treatments is likely to increase due to restricted growth in the non-cultivated treatment with weed invasion. The Cross Slot drill, while more expensive than the triple disc, was cost effective due to the greater production.

Historically sub clover has been widely used for pasture improvement in summer dry pastures in both South and North Islands. However its use has declined with more focus on white clover since the 1970's. In recent years there has been resurgence in its use due to the release of new cultivars from Australia coinciding with a series of dry summers in New Zealand in the late 1990's. These dry summers highlighted some of the weaknesses of white clover in extended drought and were the catalyst for not only the renewed interest in sub clover use but also research into production and persistence of sub clover cultivars and their grazing management. The Lincoln University Mt Grand high country station in Central Otago where this study was based is typical of farms in that area with a history of sub clover in early years of hill country pasture improvement but over the last 30 or so year's only white clover has been oversown. This project recorded where early sub clover introductions have persisted, where white clover has established and the area covered by volunteer annual clovers that make up the total clover herbage on offer to grazing animals. Four experiments were conducted both across the farm and in small plots from November 1999 to December 2003. The aim was to determine the sub clover zone using cover % to estimate botanical composition, exclusion cages to measure herbage production and temperature data loggers and rain gauges to characterize climatic conditions. Climate stations recorded air and soil temperatures at 5 sites, four on sunny aspect at 450, 620, 750, 910 m a.s.1. and one south facing site at 630 m a.s.1. The effect of altitude and aspect on temperature and rainfall was compared and thermal time (ºCd) calculated. Thermal time for the spring growth period (mid Aug-Dec) was about 1400 °C.d at 450N and declined at 100 °C.d per 100 m elevation (r² = 0.99) to be < 800°C.d at 910N. Lower accumulated thermal time with increasing altitude may restrict full lifecycles of annual clover (germination to mature seed). Rainfall differed little with altitude or aspect making temperature and associated evapotranspiration the drivers of botanical composition. Interestingly the 630S site with its southerly aspect had thermal time values between those for the 620N and 750N sites. A preliminary botanical survey showed that sub clover and volunteer annuals were the predominant clovers on offer on the sunny faces. Herbage production from exclusion cages placed on favourable sunny face sites confirmed this with 25% of November/December herbage production of 4000 kg/DM/ha coming from sub clover cultivar 'Mt Barker'. More detailed botanical composition analysis through visual cover scoring was done over two growing seasons in 2002 and 2003, next to the five climate stations. The contrast in botanical composition between north and south faces was dramatic. White clover was between 12 and 40% of cover on the south face compared to less than 5% on sunny faces with annual clovers the dominant on sunny faces. Adventive clovers cluster (T glomeratum), striated (T stratum), haresfoot (T arvense) and suckling (T dubium) made significant contributions to sunny face herbage production. Suckling clover was the only annual clover recorded on the south face. It was concluded that white clover had occupied all suitable niches and further spreading of its seed need not continue. Instead earlier flowering cultivars of sub clover should be introduced to drier steeper faces currently dominated by cluster and striated clovers.

New perennial ryegrass cultivars, improved endophyte strain and increased use of nitrogen fertiliser are three technological changes available to New Zealand pastoral farmers. Understanding their interactive effects is a priority as it is likely that in the future these will be used together. Previous studies have noted interactions between these three factors determined above and have concerns for the effect that they have on endophyte concentration in pastures. A study was conducted at the Iversen Field Research Area, Lincoln University, Canterbury, New Zealand to determine the difference in response to nitrogen fertiliser of perennial ryegrass cultivars varying in endophyte strain. Three replicated blocks were planted in a factorial combination of cultivar and endophyte with cultivars chosen to represent a comparison of standard diploid grasses with new cultivar development. Nitrogen was applied at 0 and 160 kg/N/ha. Endophyte score, Argentine stem weevil damage and pasture measurements were collected to determine the difference in response to nitrogen fertiliser. There were several key results of this trial. There was no effect of ploidy or hybrid cultivar and endophyte strain (AR1, AR37, HE) on endophyte concentration. The percentage Argentine stem weevil damage was lower in plants infected with AR1 (AR1, 18% tiller damage, Nil, 26.2% tiller damage) but was not affected by cultivar or any other interactions. All perennial ryegrasses infected with AR1 and AR37 had less clover than nil endophyte in April 2007 but in June 2007, the effect was only significant in 'Aberdart'. There were relative small effects of endophyte strain (AR1, AR37, HE) on pasture growth in January and April and in June, there, was an endophyte x cultivar interaction. There were more tillers in 'Aberdart' than any other cultivar; however, it also had the lowest tiller mass (g DM). Application of nitrogen in 'Aberdart' caused a (25%) increase in the tiller number and in 'Samson' caused a (29%) increase in tiller mass. There was an endophyte x cultivar interaction in 'Quartet' where plants infected with ARI produced more tillers/m2 than endophyte free grass .. These results may have implications for pastoral farmers. In the first field based study of the interactive effects of endophyte strain, cultivar and nitrogen there is limited evidence of reduced endophyte concentration, but there is a need to monitor over the next 2-3 to detect any further differences that may emerge. The change in clover content is important and while endophyte may increase persistence it may decrease clover and hence the feeding value of pastures. Pastures in the North Island are exposed to a wider range of insect pests than those in Canterbury. Therefore endophyte concentration might be more important in these areas of the country with these results having a more pronounced affect than in Canterbury.

The white clover (Trifolium repens) content of finely mixed ryegrass-white clover pastures is often less than 20%. As a consequence, new methods of pasture establishment and management are sought to increase white clover content. Pastures were established on 15 March 2005 at four perennial ryegrass sowing rates: 2.5, 5, 10, 20 kg/ha and using four sowing methods: standard (15 cm drill rows), spatial separation (alternating drill rows of perennial ryegrass and white clover), broadcast (seed spread on soil surface) and temporal separation (white clover sown initially, perennial ryegrass overdrilled later). White clover was sown at 3 kg/ha and chicory (Cichorium intybus) was sown at 1.5 kg/ha in all plots. Dry matter production, botanical composition, soil temperature distribution, chicory, white clover and perennial ryegrass morphology and extent of clover grazing were assessed. From January - September 2006, total dry matter production was greater in temporal treatments (4665 kg DM/ha) than spatial (3102 kg DM/ha), standard (2801 kg DM/ha) and broadcast (3300 kg DM/ha) treatments. The temporal treatment also had a higher average white clover percentage (12%) than spatial (5%), standard (5%) and broadcast (7%) treatments. Dry matter production was similar for all rye grass sowing rates, but white clover percentage was higher in 2.5 kg/ha treatments (8%) than in 5 kg/ha (7%), 10 kg/ha (7%) and 20 kg/ha (6%) treatments. White clover plants between ryegrass drill rows in the spatial sowing treatment had higher total stolon length, leaf dry weight, tap root weight and tap root diameter, and more primary and secondary stolons and total growing points than those located in ryegrass drill rows in standard and spatial sowing treatments. Moreover, there was a greater population of these plants between ryegrass drill rows in spatial plots, and chicory plants in these locations had larger basal diameters. During an autumn grazmg event, a greater proportion of white clover leaves were grazed for plants located between ryegrass drill rows in spatial treatments than in ryegrass drill rows in spatial and standard treatments. Results from this trial show that the temporal sowing method led to increased dry matter production and white clover content in pasture but there was no difference between standard, spatial or broadcast sowing methods. It appears that spatial separation, while minimising competition between ryegrass and white clover and chicory, it also allowed sheep to show greater selection for clover. The results also show that sowing rates of ryegrass are commonly too high, with no losses in production shown from lower sowing rates (e.g. less than 10 kg/ha).

Perennial ryegrass (Lolium perenne L.) pastures were sown on 19 February at four sowing rates (0, 5, 10, 20 kg/ha) with four rates of nitrogen fertiliser (0, 20, 40, 80 kgN/ha). White clover (Trifolium repens L.) was sown at 3 kg/ha in all plots. Ryegrass and white clover seedling development, herbage production, botanical composition, herbage nitrogen percentage and grazing preference of autumn sown pastures were assessed. There was no advantage in terms of dry matter production from sowing greater than 10 kg/ha. Total dry matter from at 10 kg/ha was 3600 kgDM/ha from 19/02/04 until 09/09/04. In September (203 DAS) there was no difference in ryegrass plant population numbers in the 10 and 20 kg/ha rates and sowing at 5 kg/ha gave the highest clover populations with 127 plants/m² compared with 55/m² at 20 kg/ha. Sowing at 5 kg/ha gave highest white clover content within the sward; 9.8% (24 April, 36 DAS), declining to 1.6% at 20 kg/ha and 5.7% (9 September) declining to 2.7%. There was no difference in weed suppression between 5 and 20 kg/ha sowing rates at all harvest. At 41 DAS white clover shoot and root weights were 25% greater in ryegrass pastures sown at 10 kg/ha or less. By spring shoot and root weights in the 5 and 10 kg/ha pastures were 180% and 380% greater than 20 kg/ha treatments. In September clover plants had an average of 8.5 trifoliate leaves/plant in 5 and 10 kg/ha pastures, compared with 2.6 trifoliate leaves in the 20 kg/ha pastures. Clover seedlings sown in the 0 kg/ha plots averaged 11.4 trifoliate leaves/plant. Pastures sown at 5 kg/ha and 10 kg/ha contained had an overall average nitrogen content of 4.3%, compared with 4.0% for 20 kg/ha pastures. This may have influenced grazing preference. Pastures sown at 10 kg/ha were utilised (>75%) more than 20 kg/ha pastures (

Sodium (Na) is not essential for plant growth however it is required by grazing ruminants. Natrophobic species are usually deficient for animal nutrition (

The establishment phase of six pastures was studied over a seven-month period. Four clover species plus cocksfoot were compared with white clover/ryegrass and lucerne pastures. All six pastures were sown into a cultivated fertile Templeton silt loam seedbed on February 18. Treatments were replicated 4 times, in randomised blocks, with the 0.05 ha plot individually fenced for grazing by sheep. Seedling establishment on March 6 from field counts found grass populations ranged between 380 to 230 plants m⁻² and 370 to 120 plants m⁻² for legumes, by June 19 this decreased to between 140 and 290 plant m⁻² for the grasses and between 60 and 170 plants m⁻² for the legumes. On April 17 seedling legume leaf numbers were 41.1, 30.2, 26.2, 14.5, 9.9 and 5.7 leaves plant⁻¹ for sub clover, lucerne, balansa clover, white clover (cocksfoot), white clover (ryegrass) and Caucasian clover respectively. On June 25, destructive samples found ryegrass had 13.6 and cocksfoot 7.0 tillers plant⁻¹. Cocksfoot grown with sub clover had 1.3 tillers plant⁻¹ less. Total seedling weights saw sub clover with 0.630 g plant⁻¹ , compared to 0.480, 0.350, 0.160, 0.120, 0.070 g plant⁻¹ for lucerne, balansa, Caucasian, white (cocksfoot), white (ryegrass) clovers respectively, on April 17. By September 12, dry matter botanical compositions showed that balansa clover/cocksfoot pastures had 60% and sub clover/cocksfoot, white clover/cocksfoot, white clover/ryegrass and Caucasian clover/cocksfoot pastures had 32%, 18%, 13% and 5% legume respectively. By September 8, ryegrass pastures had produced 4050, compared to 3790, 3480, 3470, 2880, 1520 kg DM ha⁻¹ for balansa clover/cocksfoot, white clover/cocksfoot, sub clover/cocksfoot, Caucasian clover/cocksfoot and lucerne pastures respectively. These results showed that annual clovers contributed more legume than perennial clovers through the cool season after a February establishment, and that ryegrass suppressed white clover growth and development more than cocksfoot. Lucerne seedlings were as productive as sub and balansa clovers in autumn but were dormant through the winter. Caucasian clover produced few leaves before winter and was suppressed by cocksfoot after the early autumn sowing.

A study was undertaken to evaluate the legume content in dryland pastures. Most dryland farmers continue to sow ryegrass/white clover pasture mixtures, but white clover fails to persist in drought conditions, leading to poor quality pastures. The objective of the study was to find out how much lucerne and subterranean clover are currently being utilised, to see which clovers are sown and the to assess the persistence of clovers, based on farmers' visual assessments of their pastures. A postal questionnaire was sent to 360 light land farmers in Canterbury and North Otago. The questionnaire focussed on pasture types, sowing rates, the use of greenfeeds, pasture management and animal carrying capacities on each property. A total of 82 questionnaire responses were received (26%), 7 were discounted due to the small size of the properties, leaving 75 for the analysis. For the analysis of the results the respondents were stratified into those growing lucerne (n=50) and those growing pastures only (n=25). From the questionnaire respondents ten were selected for in-depth interviews, five that grew larger areas of lucerne and five that had predominately subterranean clover based pastures. The interviews and pasture assessments were conducted in October and November 2000. The results showed that for the farms growing lucerne (66.7%) the average area is only 17% of the total farm area. The lambing percentage (survival to tailing) was 132% on lucerne farms and 117% on farms with 'no lucerne', but for the interviewed farmers growing lucerne the lambing percentage was 147%, with 124% for the no lucerne farms. The improved lambing percentage may be attributed to better ewe nutrition resulting in heavier liveweight ewes and the ability to flush ewes over mating. The predominant ewe breed was Border Corriedale (45% of the farms surveyed), Corriedale (24%) followed by Coop worth and Border Romney. The preferred sire was Polled Dorset (24%), Border Corriedales (22%) then Dorset Downs and Texels. The white clover cover as a percentage of total pasture cover was estimated by farmers as being 21 % in young pastures, but they estimated that this declined to 13% in older pastures. Ninety seven percent of the surveyed farmers sowed white clover in their pasture mixes at an average sowing rate of 2.35kg/ha. Farmers believed that white clover persisted in their pastures for 6.2 years on average, yet they only renewed their grass/clover pastures every 9.9 years for the lucerne farmers and 12.2 years for pasture only farmers. This implies that pastures are without white clover for between one third to one half of the lifetime of the pasture. Red clover is sown by 38% of the farmers at an average of 2.35 kg/ha, but has little contribution to pastures at less than 1 % of pasture cover on average. Only 32% of farmers sowed subterranean (sub) clover in their pasture mixes despite the fact they believed it persists for 15.4 years on average (beyond their current pasture renewal date). The farm owners visually estimated the sub clover botanical cover to constitute 11 % in young and 6% in older pastures on the surveyed farms. On the interview farms, sub clover made up 41 % of the pasture cover in young and 35% in old pastures. Meanwhile on the farms with no lucerne, sub clover was estimated to comprise 22% in young and 21 % in the older pastures of the lucerne farms, when the farms were visited in mid/late spring. The average area of winter greenfeed sown was 15 hectares (8% of farm area) for the sheep and beef farmers who did not winter graze additional stock, but 18 hectares (10%) for the sheep and beef farmers who wintered dairy cows or deer. The latter group represented 18% of all the farmers surveyed. Annual grass was the most commonly sown greenfeed type (25%), followed by turnips (18%) and then turnips and grass (14%). There was little difference in the area of greenfeed sown between lucerne and pasture only farms.