Massey Documents by Type

Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294

Browse

Filters

Settings

Subject: search.filters.subject.Weeds

Search Results

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    Deep learning-based approaches for plant disease and weed detection : a thesis by publications presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Engineering, Massey University, Auckland, New Zealand
    (Massey University, 2022) Saleem, Muhammad Hammad
    To match the ever-growing food demand, the scientific community has been actively focusing on addressing the various challenges faced by the agricultural sector. The major challenges are soil infertility, abrupt changes in climatic conditions, scarcity of water, untrained labor, emission of greenhouse gases, and many others. Moreover, plant diseases and weeds are two of the most important agricultural problems that reduce crop yield. Therefore, accurate detection of plant diseases and weeds is one of the essential operations to apply targeted and timely control measures. As a result, this can improve crop productivity, reduce the environmental effects and financial losses resulting from the excessive application of fungicide/herbicide spray on diseased plants/weeds. Among various ways of plant disease and weed detection, image-based methods are significantly effective for the interpretation of the distinct features. In recent years, image-based deep learning (DL) techniques have been reported in literature for the recognition of weeds and plant diseases. However, the full potential of DL has not yet been explored as most of the methods rely on modifications of the DL models for well-known and readily available datasets. The current studies lack in several ways, such as addressing various complex agricultural conditions, exploring several aspects of DL, and providing a systematic DL-based approach. To address these research gaps, this thesis presents various DL-based methodologies and aims to improve the mean average precision (mAP) for the identification of diseases and weeds in several plant species. The research on plant disease recognition starts with a publicly available dataset called PlantVillage and comparative analyses are conducted on various DL feature extractors, meta-architectures, and optimization algorithms. Later, new datasets are generated from various local New Zealand horticultural farms, named NZDLPlantDisease-v1 & v2. The proposed datasets consist of healthy and diseased plant organs of 13 economically important horticultural crops of New Zealand, divided into 48 classes. A performance-optimized DL model and a transfer learning-based approach are proposed for the detection of plant diseases using curated datasets. The weed identification has been performed on an open-source dataset called DeepWeeds. A two-step weed detection pipeline is presented to show the performance improvement of the deep learning model with a significant margin. The results for both agricultural tasks achieve superior performance compared to the existing method/default settings. The research outcomes elaborate the practical aspects and extended potential of DL for selected agricultural applications. Therefore, this thesis is a benchmark step for cost-effective crop protection and site-specific weed management systems (SSWM).
  • Thumbnail Image
    Item
    The potential ability of Stethorus bifidus (Kapur) to regulate populations of Tetranychus lintearius (Dufour) : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Ecology at Massey University
    (Massey University, 1993) Peterson, Paul Gordon
    Tetranychus lintearius Dufour (Acari; Tetranychidae) was released in New Zealand, in 1989, to assist in the regulation of gorse (Ulex europaeus). The present study was initiated to determine why colonies of GSM are consistently failing to establish above parallel 39°S and the possibility that Stethorus bifidus may be regulating populations of T.lintearius (GSM). Predation by S.bifidus was investigated by examining both numerical and functional responses to prey density. - Development of S.bifidus is described by a linear relationship with temperature between 8.5°C and 27.5°C (numerical response). Oviposition and temperature are linearly related and independent of GSM density. Measurements of temperature under GSM webbing showed an elevation of 1- 20 C above ambient. - The feeding rate of S.bifidus increased in a non-linear fashion between 6.5°C and 32.5°C (functional response). - Handling time decreased with increasing prey density demonstrating that S.bifidus is an effective predator at high mite densities. This investigation suggests that the role of S.bifidus in regulating GSM is more important in northern regions of New Zealand.
  • Thumbnail Image
    Item
    Responses of Rumex obtusifolius L. to several 'hormone' herbicides : being a thesis submitted in partial fulfilment of the requirements for the degree of [Master of] Agricultural Science, University of New Zealand, Massey Agricultural College
    (Massey Agricultural College, 1952) Templeton, John Keith
    Weeds have been a problem to man ever since he began to till the soil. Their presence is a factor lowering yield and increasing the cost of production of almost every economic crop. Weed eradication and control measures therefore, are bound to loom large in the management of crops at various stages throughout their growing period. Following upon such discoveries as those of Slade et al (1) and Mitchell and Hamner (2) that synthetic growth-regulating substances possessed properties capable of causing violent and often fatal disturbances to plant growth, new and effective methods of destroying undesirable species were developed. In fact, it can be said that the whole thought and practice of weed control was revolutionized by the promotion of certain practical aspects proceeding from the rapidly accumulating knowledge of plant-growth substances. [FROM INTRODUCTION]
  • Thumbnail Image
    Item
    The effect of perennial ryegrass (Lolium perenne) sowing rates on weed establishment, after one year : a thesis presented in partial fulfilment of the requirements for the degree of Master of Applied Science in Plant Science at Massey University
    (Massey University, 2001) Armstrong, Michelle Louise
    Two trials were conducted to determine whether the weed control benefits of high sowing rates of perennial ryegrass (Lolium perenne) continued into the second year after sowing. Seeds of Scotch thistle (Cirsium vulgare), nodding thistle (Carduus nutans), ragwort (Senecio jacobaea) and hedge mustard (Sisymbrium officinale) were sown into established perennial ryegrass and white clover (Trifolium repens) swards both in trays and in the field, in late autumn. The emerging weed seedlings were mapped at intervals to enable their fate to be followed. In the trays, seedlings were mapped at four week intervals, and the experiment continued for four or five months, depending on the species. In the field, weed seedlings were mapped both before and after each grazing and the experiment continued for seven months. The trays with low ryegrass sowing rates had a higher density of white clover than those trays with high ryegrass sowing rates. In the field plots with low sowing rates, both the broad-leaved weed population and volunteer grasses were more abundant than in the high ryegrass sowing rate plots. This resulted in the different treatments having a similar competitive ability against the sown weed seeds, one year after the swards were sown. Both nodding thistle and Scotch thistle emergence was extremely low in the bare ground trays, probably relating to seed predation of these large seeds off bare soil. Scotch thistle and hedge mustard emergence were both reduced by high sowing rates of perennial ryegrass when compared to trays and field plots sown with white clover only. Nodding thistle and ragwort however provided some unusual results. Both of these species experienced an increase in total emergence as ryegrass sowing rate increased. With nodding thistle, this occurred in the field and with ragwort, in the trays. This obviously was unexpected but in reality the increases in emergence, although significant, were small and of little practical importance. The increases probably related to the similar competitive ability of the swards one year after pasture sowing. Emergence of ragwort in the field and nodding thistle in the trays was not affected by ryegrass sowing rate. For all species, in both the field and the trays, emergence was delayed as the sowing rate increased. There was very low mortality in the tray experiment. Mortality was much higher in the field due to the effects of grazing animals and also the longer time frame of the field experiment. All hedge mustard seedlings that emerged in the field were killed, and only one nodding thistle seedling survived to the end of the experiment. With both Scotch thistle and ragwort, seedlings seemed to have a higher chance of surviving in the high density ryegrass treatments, probably due to the physical protection from grazing animals provided by ryegrass. High mortality in the field experiment rendered seedling size data meaningless. In the trays, for all species, the heaviest seedlings were found in the treatments with either bare ground or where only clover was sown, indicating that competition from ryegrass had a serious negative effect on weed size. Weed size did not vary significantly as ryegrass sowing rate increased. In conclusion it would appear that most of the weed control advantages from higher sowing rates of perennial ryegrass are obtained in the first year after sowing, as all swards that contain perennial ryegrass appear to have a similar competitive ability against invading weeds in the second year after sowing.
  • Thumbnail Image
    Item
    Invasion of woody species into weed infested areas : a thesis presented in partial fulfillment of the requirements for the degree of Masters of Science in Ecology at Massey University
    (Massey University, 2000) Trass, Amy Patricia
    When studying plant dynamics and succession, it is important to determine potential limiting factors affecting recruitment (Crawley 1990). The purpose of this study was to investigate factors affecting the establishment and survival of woody species in weed infested areas around the central volcanic plateau. This was achieved by first describing these communities, and quantifying the number of native seedlings and saplings found in both forested and non-forested (weedy) areas. Seed input was measured with seed traps, and factors affecting recruitment of seedlings were investigated by manipulative field experiments. Although some native woody species were dispersed into weedy areas, both seed and seedling densities of most species declined rapidly with increasing distance from the forest margin. Sowing seeds at densities equivalent to 625 per m significantly increased seedling establishment of Griselinia littoralis and Coprosma 'taylorii' but not Pittosporum tenuifolium var. colensoi. Removal of exotic grasses (clearing treatment) that dominated in non-forest areas resulted in much greater establishment of all woody seedlings, including introduced broom (Cytisus scoparius) and several native species that had dispersed naturally. Most species also showed greater establishment in plots that were caged to prevent predation. However, the effects of clearing and caging treatments on survival of seedlings were not as apparent as they were for establishment. In addition, experimental clearing increased the growth of transplanted G. littoralis seedlings. Overall, most native species had much lower seedling establishment and survival in weedy areas compared with native forest. This is explained by a combination of both seed and microsite limitation in weedy areas. In another experiment designed to test the effects of bird consumption on seed germination, bird dispersers increased germination percentages of native species by removing fruits from seeds. All species examined (G. littoralis, Coprosma robusta, Pseudopanax crassifolius, and P. tenuifolium var. colensoi) showed very low germination of seeds within fruit, and much greater germination of seeds that were cleaned either by passage through birds or by hand. For C. robusta, G. littoralis, and P. tenuifolium var. colensoi, passage through birds also significantly increased germination of seeds compared with those cleaned by hand. The rate of germination was less affected by different treatments than the absolute percentage germinating, but was generally faster in bird-voided compared to hand-cleaned seed. Seeds in both of these latter treatments germinated considerably faster than seeds within fruit. Invasion of native woody plants in weedy areas appears to be constrained by a combination of low rates of seed dispersal for most species, and low probabilities of seedling establishment and survival in areas without disturbance. The most likely future scenario for the majority of weedy areas studied is continued dominance of exotic species in the short term, with slow succession to native shrubland as well-dispersed, frost-resistant native species such as manuka (Leptospermum scoparium) establish after disturbance. Management options are discussed with the aim of accelerating the rate of succession in weedy areas to native forest.
  • Thumbnail Image
    Item
    A study of the germination and establishment of lucerne (Medicago sativa L.) oversown on an uncultivated grass sward : a thesis presented in partial fulfillment of the requirements for the degree of Master of Agricultural Science (M. Agr. Sci.) at Massey University
    (Massey University, 1977) Kusekwa, Martin Lawrence
    In this study on the oversowing of lucerne on an uncultivated sward at Massey, three trials were conducted, two in the field and one box trial under glasshouse conditions during the summer of 1975/76 and the autumn and early winter of 1976. The triars compared the effects of two lengths of ground cover, untreated, suppressed for a longer or shorter period with herbicides and burnt on the early establishment of lucerrne sown as bare or pelleted seed. It was found that small amounts of cover at sowing encouraged lucerne germination under high moisture conditions subsequent establishment and early survival was best on the burnt and chemically desiccated, plots. The cover was most useful where it was short (8 cm) and desiccated by long term paraquat or glyphosate treatments. Dense live cover was harmful to lucerne seedlings and smothered them within 5 weeks after sowing. Where recovery of the resident vegetation and weeds was rapid, as on the burnt and short term herbicide plots, lucerne establishnent and survivar was poor and surviviag plants were small and weak. Weed infestation was heavy both on burnt and herbicide treated plots, pointing to the need to introduce an early weed control programme with oversowing. Weeds were found to respond in a similar manner to lucerne to grass competition. In the autumn and early winter tria1s, slug damage was observed and this problem appeared to be of major importance under wet conditions. The effects of slugs were fourd to be most damaging during the first 4-5 weeks after sowing. Metaldehyde slug bait pellets were effective in reducing slug damage if applied at least twice during the establishment period of lucerne. The implications of these findings are reported and discussed with particular reference to the possibilities of oversowing under East African conditions
  • Thumbnail Image
    Item
    Integration of goats into sheep and cattle grazing systems as a permanent weed control tool : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Plant Science, Institute of Natural Resources, Massey University, Palmerston North, New Zealand
    (Massey University, 2001) Beskow, Wagner Brod
    This project evaluated the integration of goats into sheep and cattle grazing systems as a permanent herbaceous weed control tool. Foraging behaviour of goats grazing pasture with no access to browse was monitored in Experiment 1. A selection index was used to determine degrees of effort for selection or avoidance of plant species. Although most of the grazing time was spent on herbage species, no effort was made to select them, whereas moderate to strong efforts were made to select weeds. Selection index varied with season of the year. Goats preferred taller vegetation and grazed above 10 cm from the ground whenever possible, though they took 70% of their bites from below the top of the sward, regardless of organ selected. Experiment 2 consisted of feeding known numbers of weed and herbage seeds to goats and recovering them to assess destruction and viability. On average, 92% of the viable seeds were destroyed, with up to 72 hours for complete excretion. The effects of goat and cattle grazing on the structure and competitiveness of swards against weed invasions were studied in Experiment 3. A pasture competitiveness index (PCI) was calculated using structural attributes of a sward resulting from goat and cattle grazing alone and in combinations, under hard and lax grazing severities. The sward was denser, leafier, more uniform, and covered the soil more thoroughly (greater PCI) with increasing participation of goats in the treatments, especially because of severe treading damage caused by cattle contrasting with no damage from goats. The PCI explained 93% of the variation in weed seedling density. The effect of pasture surrounding thistle rosettes on their consumption by goats was investigated in Experiment 4. Thistle rosettes that were exposed by clipping the adjacent sward were severely defoliated by goats, while rosettes surrounded by the sward were only nibbled or not eaten at all. Grazing management, economics, and drawbacks of using goats for herbaceous weed control are discussed. Integration of goats into grazing systems can result in a more desirable botanical composition, more uniform and competitive sward, reduced weed control costs, greater herbage utilisation, and better pasture quality.