Author
![]() |
PETHYBRIDGE, SARAH - Botanical Resources Australia Pty Ltd |
![]() |
HAY, FRANK - Tasmanian Institute Of Agricultural Research |
![]() |
Gent, David - Dave |
Submitted to: Plant Disease
Publication Type: Peer Reviewed Journal Publication Acceptance Date: 6/22/2010 Publication Date: 11/1/2010 Citation: Pethybridge, S.J., Hay, F.S., Gent, D.H. 2010. Characterization of the spatiotemporal attributes of Sclerotinia flower blight epidemics in a perennial pyrethrum pathosystem. Plant Disease. 94:1305-1313. Interpretive Summary: Sclerotinia flower blight causes substantial crop losses in pyrethrum, the source of the naturally occurring pyrethrin insecticdes. We studied the dynamics of disease development in space and time to better understanding how outbreaks occur and how they may be mitigated. Various statistical tests indicated that appearance of the pathogen is closely linked to flower development in the host, with very patchy occurrences of diseased flowers. Disease epidemics appear to be driven by sources of the pathogen very close to the individual plants, and strategies that reduce the amount of overwintering inoculum may be the most efficient way to manage this disease. Technical Abstract: Sclerotinia flower blight, caused by Sclerotinia sclerotiorum causes substantial direct crop losses from reductions in the numbers of harvestable flowers in Australian pyrethrum fields. The pathogen can also cause plant death from crown rot through myceliogenic germination. The spatiotemporal characteristics of Sclerotinia flower blight epidemics were characterized in five fields over a three year period (2007 to 2009). In all fields, the incidence of Sclerotinia flower blight was assessed at least five times throughout the flowering period. At each assessment, apothecia counts along the predefined transects were also conducted. Spatial analyses of the epidemics were characterized by fitting the beta-binomial and binomial distributions, the binary power law, spatial autocorrelation using first and second order statistics, and ordinary and median runs analysis. Log likelihood tests indicated that the beta-binomial distribution fit better than the binomial distribution for 92% of the data sets. The index of dispersion, D, was significantly greater than 1 in 97% of the data sets, with a median value of 5.58. The estimated parameters of the slope and intercept terms of the binary power law were 1.631 (SE = 0.059) and 0.678 (SE = 0.099), indicating a higher degree of aggregation at the scale of individual sampling units. In 69% of the data sets, the magnitude of the first order autocorrelation coefficient, , was significantly greater than 0 (indicating aggregation), with median value of 0.41. Results of these analyses combined suggested a high degree of aggregation of disease incidence at the scale of the sampling unit, with patches of similar disease status often extending to approximately 1 m. Temporal analysis of epidemics was conducted using nonlinear regression and back-transformation to select the model of best fit. In 11 of the 12 epidemics, the monomolecular model was found to provide the best fit to the data, indicative of monocyclic processes. A significant spatial association between apothecia and incidence of Sclerotinia flower blight within the lag of one sampling unit was quantified using Spearman’s rank cross-correlation coefficients, suggesting very localized dispersal of ascospores in pyrethrum fields. Taken together with the results of the temporal analysis, the current study suggests the S. sclerotiorum apothecia emergence is closely synchronized with pyrethrum flower development and epidemics appear to be dominated by localized sources of ascosporic inoculums due to limited escape from the pyrethrum canopy. This research will provide the basis for the design and implementation of improved management strategies for this disease in pyrethrum fields. |