Teses e Dissertações
URI permanente desta comunidadehttps://locus.ufv.br/handle/123456789/1
Teses e dissertações defendidas no contexto dos programas de pós graduação da Instituição.
Navegar
2 resultados
Resultados da Pesquisa
Item Disentangling environmental effects on plant disease epidemics at the regional scale(Universidade Federal de Viçosa, 2022-11-17) Alves, Kaique dos Santos; Del Ponte, Emerson Medeiros; http://lattes.cnpq.br/3166163630863998The environment plays an essential role in driving the occurrence and dynamics of plant disease epidemics. The weather is known to influence pathogen and host biology and should modulate the stages of the disease cycle. On the other hand, climate, the average of weather within long periods of time (i.e. 30 years), should set the predominance of pathogens and host genotypes across regions, and therefore the spatial distribution of plant diseases and their respective intensities. In this thesis, three studies aiming to unravel the effects of environmental factors on plant disease epidemics will be presented. The first study will demonstrate how the climate shapes the spatial distribution of citrus Huanglongbing prevalence in Minas Gerais. In the second study, the time to onset of soybean rust in commercial soybean fields from Southern Brazil was associated with the El Nino Southern Oscillation, a phenomenon that triggers extreme weather events around the globe and also in Brazil; The third study integrates cutting- edge statistical methodology to associate weather time series and soil properties data to white mold prevalence in snap bean fields in New York, United States. The results led to novel insights into pathogen biology and disease risk at the regional and local scales for the three pathosystems under study. Keywords: Epidemiology. Huanglongbing. Soybean rust. White mold. Bayesian. Machine learning.Item Estimation of the time-varying apparent infection rate from plant disease progress curves: a particle filter approach(Universidade Federal de Viçosa, 2019-08-12) Alves, Kaique dos Santos; Del Ponte, Emerson Medeiros; http://lattes.cnpq.br/3166163630863998The parameters of the simplest (two-parameter) epidemiological models that best fit plant disease progress curve (DPC) data are biologically meaningful: one is the surrogate for initial inoculum (𝑦 0 ) and the other is the (constant) apparent infection rate (𝑟), both being useful for understanding, predicting and comparing epidemics. The assumption that 𝑟 is constant is not reasonable and fluctuations are expected due to systematic changes in factors affecting infection (e.g. weather, host susceptibility, etc.), thus leading to a time-varying 𝑟, or 𝑟 𝑘 , being 𝑘 = 1,2, . . . , 𝑁 and 𝑁the final epidemic time. A rearrangement in formulation of these models (e.g. logistic, monomolecular, etc.) can be used to obtain 𝑟 between two time points, given the disease (𝑦) data are available. We evaluated one of the several data assimilation techniques, the Particle Filter (PF), as an alternative method for estimating 𝑟 𝑘 . Synthetic DPC data for hypothetical polycyclic epidemics were simulated using the logistic differential equation for scenarios that combined five patterns of 𝑟 𝑘 (constant, increasing, decreasing, random or sinusoidal); five increasing time assessment interval (𝛥𝑡 = 1, 3, 5, 7 or 9 time units - t.u.); and two levels of noise (0.1 or 0.25) assigned to 𝑦 𝑘 . The analyses of 50 simulated 60-t.u. DPCs showed that the errors of PF-derived𝑟̂ 𝑘 were lower (RMSE < 0.05) for 𝛥𝑡 < 5 t.u. and less affected by the presence of noise in the measure compared with the logit-derived 𝑟 𝑘 . The ability to more accurately estimate 𝑟 𝑘 may be useful to increase knowledge of field epidemics and identify within-season drivers of 𝑟 𝑘 behaviour. Keywords: Data assimilation. Inverse problems. Sequential Monte Carlo.