Adam H. Sparks
Research Associate,
KSU Department of Plant Pathology
asparks@ksu.edu ·
785.532.1341
Adam Sparks is a research associate at Kansas State University in the Department of Plant Pathology at Kansas State University in Manhattan, Kansas. who works with Dr. Erick De Wolf after completing his PhD with Dr. Karen Garrett's Plant Disease Ecology Lab. He spends most of his days staring at two computer monitors and attempting to write computer code to model plant disease and reading the occasional scientific article.
His interests and expertise include epidemiology and the ecology of plant diseases; GIS applications in plant pathology; agricultural extension outreach and teaching. His Ph.D. studies focused on major plant pathogen groups, ecology and epidemiology, modeling, GIS, and remote sensing. His full CV is available for download as a pdf file, Adam Sparks's Curriculum Vitae.
His PhD dissertation research involved scaling plot-level models up to larger spatial and temporal scales. Many plant disease forecasting models are created using small time-step (e.g., hourly or less) weather data to predict disease severity or the need for control measures to be enacted. Several sources of large time-step (e.g., daily or monthly) weather summaries exist in ready to use datasets. In certain situations rather than backcasting to create hourly weather data it may be desirable to create a quick estimation of disease severity using these coarser datasets.
Dissertation
"Disease risk mapping with metamodels for coarse
resolution predictors: Global potato late blight risk now and
under future climate conditions"
Pathosystem: late
blight of potato caused by Phytophthora infestans
Initial Model Used for Development: SimCast
using hourly time-step data, written in R
Metamodels for large time-step weather data:
Generalized Additive Models (GAM), written in R based on
SimCast, capable of using daily, weekly or monthly time step
weather data.
Research Objectives
- Created late blight prediction models capable of utilizing daily, weekly, or monthly time-step weather data.
- Compared models for different time-steps.
- Evaluated the effects of possible climate change scenarios on late blight severity.
- Compared effects of the use of late blight resistant cultivars with susceptible cultivars.
Web-based teaching modules for plant pathology applications in the R programming environment
These teaching modules are housed permanently on the APSNet Education Center website, Web-based teaching modules for plant pathology applications in the R programming environment