LONG-DISTANCE DISPERSAL AND DISEASE OUTBREAKS: EFFECTS OF INITIAL PREVALENCE, BASIC REPRODUCTION NUMBER, AND CONTROL TACTICS

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: TERMINATED
• Funding Source: AFRI COMPETITIVE GRANT
• Reporting Frequency: Annual
• Accession No.: 1007157
• Grant No.: 2015-67013-23818
• Project No.: ORE00131
• Proposal No.: 2015-08284
• Program Code: A1222
• Project Start Date: Sep 1, 2015
• Project End Date: Aug 31, 2021
• Grant Year: 2015

Project Director
Mundt, C. C.

Recipient Organization
OREGON STATE UNIVERSITY
(N/A)
CORVALLIS,OR 97331

Performing Department
Ag Botany & Plant Pathgy

Non Technical Summary
Empirical data and modeling studies of diseases caused by pathogens with long-distance dispersal ability will be used to: 1) Determine effects of initial disease prevalence, spatial pattern of initial disease prevalence, and pathogen reproductive capacity on disease spread; 2) Compare the efficacy of reactive ring culling, reactive ring vaccination or chemotherapeutic applications, timing and extent of reactive ringtreatments, and broad-scale population protection for disease control; and 3) Determine the influence of initial disease prevalence and pathogen reproductive capacity on the efficacy of these control tactics. Modeling studies of wheat stripe rust, foot-and-mouth disease, sudden oak death, and arbovirusesof animals will be conducted. Extensive comparative modeling will be conducted through factorial combinations of models and input data among the different diseases. Generalized theory and models will be developed to predict "rules-of-thumb" for the control of diseases caused by pathogens with long-distance dispersal. Data from field observations of sudden oak death and foot-and-mouth disease, and manipulative experiments with wheat stripe rust, will be used for model validation/verification. The project will determine the importance of initial disease prevalence, the spatial pattern of initial disease prevalence, and pathogen reproductive capacity on the spread of diseases caused by pathogens with "fat-tailed" dispersal kernels, and the interaction of these biological variables with control practices such as reactive culling, reactive vaccination or chemotherapeutic applications, and broad-scale protective strategies. The work is fundamental to our understanding of disease spread, and is crucial to predicting the spread of epidemic invasions and designing disease control strategies. The work is potentially transformative as it will provide a rare opportunity to test such hypotheses in natural and controlled field experiments, and because the applicability of a broad diversity of plant, animal, and human pathogens with fat-tailed dispersal kernels will be rigorously evaluated via the interdisciplinary modeling efforts. Conclusions should apply over a very wide range of spatial scale due to the nature of dispersal kernels of pathogens that have the potential for long-distance dispersal. The proposed work has significant potential to increase our knowledge of the spread of pathogens of taxonomically divergent hosts and pathogens and may lead to improved control of major diseases and to mitigation of invasive epidemics. Pathogens with long distance dispersal capability are responsible for some of the most potentially damaging and fastest spreading diseases of humans, livestock, and plants. The work also will make a significant contribution towards integrating work on disease spread in plant and animal systems.

Animal Health Component

50%

Research Effort Categories

Basic

90%

Applied

10%

Developmental

0%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
212	0120	1070	50%
311	0120	1070	50%

Knowledge Area
212 - Pathogens and Nematodes Affecting Plants; 311 - Animal Diseases;

Subject Of Investigation
0120 - Land;

Field Of Science
1070 - Ecology;

Keywords

long distance

disease spread

disease control

wheat stripe rust

sudden oak death

foot and mouth disease

arboviruses

epidemiology

methomatical models

Goals / Objectives
Goal: Determine the importance of initial disease prevalence, the spatial pattern of initial disease prevalence, and the basic reproduction number on the spread of diseases caused by pathogens with "fat-tailed" dispersal kernels, and the interaction of these biological variables with control practices such as reactive culling, reactive vaccination or chemotherapeutic applications, and broad-scale protective strategies.1) Use field and modeling studies to determine effects of initial disease prevalence, spatial pattern of initial diseaseprevalence, and basic infection number on disease spread of wheat stripe rust, foot-and-mouth disease, sudden oak death, and arboviruses.2) Compare the efficacy of reactive ringculling, reactive ring vaccination or chemotherapeutic applications, timing and extent of reactive ringtreatments, and broad-scale population protection for disease control.3) Determine the influence ofinitial disease prevalence and basic infection number on the efficacy ofcontrol tactics.4) Conduct extensive comparative modeling through factorial combinations of modelsand input data among the different diseases.5) Develop generalized theory and models to predict"rules-of-thumb" for the control of diseases caused by pathogens with long-distance dispersal.6) Evaluate data fromnatural experiments with sudden oak death and foot-and-mouth disease, and manipulative experimentswith wheat stripe rustfor model validation/verification.

Project Methods
Project goals will be addressed with a set of hosts and pathogens of highly divergent taxonomy, but which share the common characteristic of potential for LDD. These model systems are wheat stripe rust, sudden oak death, foot and mouth disease, and arboviruses of animals. The model systems were carefully chosen based on a history of strong prior work, high quality data for parameterizing models, and extensive previous modeling efforts. Further, in two of these systems, observational tests of hypotheses from natural experiments can be made utilizing extensive data collected during control campaigns. One of the systems will also be used in manipulative experiments to test impacts of epidemiological variables and control practices and provide validation for the modeling efforts.Hypotheses:1. Disease spread from an outbreak focus will be influenced in a predicable manner and to a greater degree by the basic production number (R0 ) than by initial disease prevalence in the focus (P0). a) Spread from the outbreak focus will increase as a power of R0. b) Spread from the outbreak focus will be approximately proportional to the square root of P0.2. The area over which P0 is distributed (focus area) will not substantially affect subsequent spread.3. Effectiveness of ring-culling and ring-vaccination (= ring fungicide in case of plant pathogens) for disease control will increase with lower values of both P0 and R0.4. Effectiveness of ring-culling and ring-vaccination will be improved in the order: earlier implementation > reduced R0 > reduced P0 > increased ring diameter. a) ring diameter will have little effect in severe epidemics, but may be important in less severe ones.5. Suppression of epidemic spread will be in the order of protection > ring culling > ring vaccination (or fungicide).6. Combining reactive ring treatments with protection of the remaining population will be more effective than reactive treatments or protection alone.7. Relationships among R0, P0, and control practices will be similar for LDD pathogens of highly diverse biology owing to similar dispersal kernels.The above hypotheses will be addressed using a combination of 1) Modeling individual diseases; 2) Observational tests (natural experiments); 3) Manipulative experiments; 4) Factorial comparisons of models and input data among disease systems and 5) Development of generalized, theoretical models to provide rules-of-thumb that are valid across a diversity of LDD pathogens.

Progress 09/01/15 to 08/31/21

Outputs
Target Audience:1) Scientists who participated in this study, via all-investigator meetings in Corvallis, Oregon, May 2016 and May 2018, and in a combination in-person/online meeting in 2017. 2) Faculty, postdocs, students, and other scientists who heard project talks presented at the following meetings: Annual Meeting of the American Phytopathological Society, 2016 US International Association of Landscape Ecology Annual Meeting, 2016 (four presentations) Annual Meeting of the Ecological Society of America, 2017 (two presentations). Institute of Electrical and Electronics Engineers Conference, 2017 9th International Congress on Environmental Modelling and Software, 2018 American Mosquito Control Association 85th annual meeting, 2018 (two presentations) Southern Forestry GIS Conference, 2018 US-International Association for Landscape Ecology, 2018 US-International Association for Landscape Ecology, 2019 (three presentations) 7th Sudden Oak Death Symposium, San Francisco, 2019 Geospatial Forum, Raleigh, NC. American Control Conference, 2019 Electrical and Computer Engineering Departments Heads Association, 2019 Ecological Society of America, 2020 3) Faculty, postdocs, and students who heard the following project seminars: University of Florida, 2017 Cornell University, 2017 Kansas State University, 2018 University of Georgia, 2018 Kansas State University, 2021 (two seminars) 4) Scientists, students, and stakeholders who have read scientific publications resulting from the project over a six-year period. 5) Graduate students who were exposed to the work in the Oregon State University course entitled "Plant Disease Management", winter 2016, 2018, and 2020 6) Graduate students who were exposed to the work in the Oregon State University course entitled "Plant Disease Dynamics", winter 2017 and 2021 7) Students who heard the following guest lectures at NC State University in three classes in 2019: Grand Challenges of Geospatial Modeling Food Security and Emerging Diseases Geospatial Computation and Simulation 8) Wheat producers who were exposed to the goals of the project over a six-year period. 9) Policymakers, California Oak Mortality Task Force, and other stakeholders who work with SOD. 10) American Mosquito Control Association members and other practitioners who work with WNV and other insect-vectored diseases. 11) Scientists and land managers who attended the Tangible Landscape Sudden Oak Death Management Workshop hosted by researchers at North Carolina State University and University of California, Davis and conducted in Salem, OR on October 10, 2017. A new geospatial tool was used to evaluate how sudden oak death management actions affect disease spread and impacts across the landscape 12) All scientists who are members of this projects, via inter-institutional visits and monthly Zoom electronic meetings in 2019-2021. This includes a two day visit of Mike Tildesely (co-PI) to Oregon State University, a four-day visit of Chris Jones (post-doc at NCSU) to Oregon State University, includes a two day visit of Oregon State University to Kansas to learn about their models, monthly Zoom meetings, and many informal Zoom meetings and face-face-meetings at each institution prior to COVID-19. 13) Land managers and other scientists who used the tangible landscape model to study epidemic mitigations. 14) Scientists and stakeholder who participated in our second participatory modeling workshop for SOD on September 10th, 2019 at the Oregon Department of Forestry in Salem, Oregon, in conjunction with the Core Science Team. 15) All scientists who are members of this project, via inter-institutional visits and monthly Zoom electronic meetings. This included a two day visit of Mike Tildsely (co-PI at University of Warwick) to Oregon State University, a four-day visit of Chris Jones (post-doc at NC State) to Oregon State University, monthly Zoom meetings, and many informal Zoom meetings and face-face-meetings at each institution prior to COVD-19. Changes/Problems:1) Postdocs often obtained jobs and left before their portion of the work was completed. Impacts were mitigated through a one-year no-cost extension to complete studies. 2) A change of naturally occurring stripe rust races had potential to interfere with wheat stripe rust experiments. This was mitigated by changing experimental sites, application of fungicides at appropriate times, and additional simulation modeling. What opportunities for training and professional development has the project provided?1) Training in modeling, experimental design, and data management was provided to six postdocs, four graduate students, and three technicians. 2) Three in-person investigator meetings and multiple Zoom meetings provided an opportunity for all senior personnel, six postdocs, four graduate students, and three technicians to interact in a multidisciplinary gathering. 3) Six postdocs, four graduate students, 3 technicians received significant experience/training in writing of scientific publications. 4) ) One postdoc and one technician received experience in international collaboration through a visit to the UK. 5) Postdocs, graduate students, and technicians gain experience presenting 21 talks at national/international meeting and 6 departmental seminars. 6) Six postdocs, four graduate students, and three technicians gained programming experience in languages such as Python, MatLab, SAS, and Pascal. 7) Four graduate students received their degrees fully or partly funded under this project. How have the results been disseminated to communities of interest?Results have been distributed to communities of interest in the following ways: 1) Publication of 28 articles in peer-reviewed journals, 4 in process of publication, and 5 in preparation 2) Publication of two Ph.D. and one Master's thesis 3) A total of 21 presentations at national or international scientific meetings 4) Presentation of six departmental seminars 5) Lectures given in seven different university classes 6) Presentation of the tangible landscape model for SOD management at three different Sudden Oak Death Workshops in Oregon 8) Wheat producers who were exposed to the goals of the project at field days over a six-year period 9) Interactions with policymakers, California Oak Mortality Task Force, and other stakeholders who work with SOD 10) Interactions with American Mosquito Control Association members and other practitioners who work with WNV and other insect-vectored diseases 11) Through exchange of visits between the U.S. and UK partners 11) Through three in-person meetings and numerous Zoom meetings among project personnel over the course of the project What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? 1) Stochastic models were used to study the spread of foot-and-mouth (FMD) disease using actual farm number and farm sizes for the states of Oregon, Pennsylvania, and Texas as well as more balanced, theoretical data sets. In these simulations, increasing cattle density always resulted in larger epidemics, regardless of farm density. Alternatively, increasing farm density only led to larger epidemics in scenarios of high cattle density. 2) We evaluated a farm-density-based ring culling strategy to control foot-and-mouth FMD in the UK. This strategy may allow for some farms within rings around infected premises (IPs) to escape depopulation, with the aim to prevent over-culling during outbreaks. 3) A stochastic model was used to estimate the epidemiological connectivity of individual farms and counties for FMD based on a combination of farm demography and spatial arrangement of farms. 4) Large field studies were established to determine effects of initial disease prevalence, spatial pattern of initial disease prevalence, and basic infection number on spread of wheat stripe rust. 5) Field studies with wheat stripe rust were conducted to determine effects of initial disease prevalence and basic infection number on the efficacy of culling for epidemic suppression. We found that culling had similar effects regardless of initial disease prevalence and basic infection number. This is consistent with simulation studies that we conducted. 6) Three biological data sets collected at very different spatial scales were computationally combined for prediction of the dispersal of cereal rusts. 7) Models based on wheat stripe rust showed that initial disease prevalence at the outbreak focus may play a larger role in subsequent disease spread than was originally anticipated by many. 8) Field experiments were conducted in two locations x two years to determine effects of the timing and extensiveness of culling and population protection of the spread and control of wheat stripe rust. Results showed that cull timing was more important than cull size and mass fungicide application had the largest impact. Simulations showed that the relationships found in the field experiments hold over a much broader range of spatial and temporal variation in mitigation treatments. 9) Effects of host density on epidemic progression and efficacy of control was investigated for wheat stripe rust. Spatially explicit, stochastic simulations indicate that epidemic interventions are much more effective in low density host populations than in high density populations. 10) Using simulations of wheat stripe rust as model system, we studied the epidemiological mechanisms to explain the extreme importance of the timing of mitigation strategies on their success that have been uncovered by our work. The work was later expanded to cover a diversity of diseases. 11) Studies of dispersal kernels have shown kernels for wheat stripe rust and foot-and-mouth disease to be remarkably similar, despite large differences in the biology of these diseases. 12) We developed a tangible landscape model for sudden oak death (SOD) spread and compared control strategies. The model shows real time epidemic outcomes, allows users to pause and rewind simulations, and compare control costs and efficacy in real time (i.e. users can adaptively move through the simulation). 13) We developed an interface that uses Tangible Landscape technology with a touchable, interactive surface to run PoPS on a local computer, while the other is a web platform that runs PoPS in the cloud. The web platform and Tangible Landscape both run the PoPS model, a sophisticated computer simulation that uses information about existing infections, locations of host trees and weather conditions to predict the spread of sudden oak death disease. We simulated the spread of SOD in California and Oregon and conducted management workshops to help stakeholders determine the most effective strategies for managing SOD in the field by comparing what-if scenarios using PoPS tangible landscape. 14) We estimated the swine movement network at farm level in the US. We found that high in-degree farm operations (i.e., markets) play critical roles in the disease spread. We also found that high in-degree based targeted isolation and hypothetical vaccinations are more effective for disease control compared to other centrality-based mitigation strategies. 15) Project participants at Kansas State University constructed a large number of models, mostly network models, to study the spread of a broad range of human and animal diseases. Goals of the models were to identify causal factors that could be potentially be managed for disease control. These models include: a. An individual-level network model for a hypothetical outbreak of Japanese Encephalitis (JE) b. A spatial metapopulation network model for West Nile Virus c. A combined network model with human and mosquito transmitted Zika virus d. An individual based heterogeneous network model to understand the spreading pattern of West Nile Virus (WNV) in the USA e. A novel, individual-based interconnected network model that incorporates both insect-vectored and sexual transmission of the Zika virus. f. Evaluated three filtering approaches to estimate Japanese encephalitis transmission rate and for a forecasting of the disease outbreak g. A generalized group-based epidemic model (GgroupEM) framework for any compartmental epidemic model h. A risk assessment framework for Ebola using a two-layer temporal network i. An individual-level network model for RVF in Uganda j. An ensemble Kalman filter (EnKF) that provides dual state-parameter estimates for the transmission of Japanese Encephalitis (JE) - a vector-borne disease k. A state-space model to model dynamic data l. Vectorial capacity was modeled and assessed for risk of vector-borne disease spread using a spatiotemporal network model and climate data with an application of dengue in Bangladesh m. A generalizable model for forecasting the spread of plant diseases with interactive management n. A mean-field group-based epidemic spreading framework was developed to reduce the computational time of our individual based Generalized Epidemic Modeling Framework (GEMF) framework o. GEMF network models were applied to Ebola and Africa swine fever transmission p. A numerical stochastic group-based epidemic spreading simulator to reduce the computational time of our individual based GEMF framework q. A group-based continuous-time Markov general epidemic modeling (GroupGEM) framework for any compartmental epidemic model r. A network model was used to study the spread and mitigation of COVID-19 in Manhattan, Kansas using the spatial distribution of actual household data 13) A project was conducted to compare four different spatiotemporal models, originally developed to model different diseases, to determine the degree to which they replicate effects of mitigation practices quantified in field experiments of wheat stripe rust. 14) Two very large modeling studies were conducted as capstones for the overall project, and with input from all project participants. The first was a full factorial of of several latent periods, basic infection rates, initial disease levels, timings of epidemic intervention, and spatial extents of epidemic intervention. Inputs were chosen to provide a balanced set representing a wide diversity of disease systems and epidemiological conditions. Results highlight the importance of both latent period and time of intervention on ability to control diseases. The second study used a reduced factorial set of the above variables, but included in the factorial an extensive set of host densities and spatial patterns. This study utilized millions of simulations, and results are still under evaluation.

Publications

• Type: Journal Articles Status: Published Year Published: 2021 Citation: Moon, S. A., F. D. Sahneh, and C. Scoglio 2021. Group-based general epidemic modeling for spreading processes on networks: Groupgem. IEEE Transactions on Network Science and Engineering 8(1):434446.
• Type: Journal Articles Status: Published Year Published: 2021 Citation: Riad, MH; Cohnstaedt, L; Scoglio, CM. 2021. Risk assessment of dengue transmission in Bangladesh using a spatiotemporal network model and climate data. Am. J. Trop. Med. Hyg. 104(4): 14441455.
• Type: Journal Articles Status: Published Year Published: 2021 Citation: Gaydos, D. A., Jones, C.M., Jones, S.K., Millar, G.C., Petras, V., Petrasova, A., Mitasova, H., and Meenetemyer, R.K. 2021. Ecol. Appl. e02446. doi:10.1002/eap.2446.
• Type: Journal Articles Status: Published Year Published: 2021 Citation: Jones, C. M., S. Jones, A. Petrasova, V. Petras, D. Gaydos, M. M. Skrip, Y. Takeuchi, K. Bigsby, and R. Meentemeyer. 2021. Iteratively forecasting biological invasions with PoPS and a little help from our friends. Frontiers in Ecology and the Environment 19(7): 411 418.
• Type: Theses/Dissertations Status: Accepted Year Published: 2021 Citation: Moon, S.A. 2021. Modeling And Analysis of Epidemic Processes Over Large Networks From Limited Data Ph.D. Thesis, Kansas State University, Manhattan.
• Type: Theses/Dissertations Status: Accepted Year Published: 2021 Citation: Ferdousi, T. 2021. Computational Models And Tools For Analysis, Prediction, and Control of Infectious Diseases. Ph.D. Thesis, Kansas State University, Manhattan.
• Type: Journal Articles Status: Other Year Published: 2022 Citation: Severns, P.M., and Mundt, C.C. 202_. Delays in epidemic outbreak control cost disproportionately large treatment footprints to offset. Sci Rep:in revision.
• Type: Journal Articles Status: Other Year Published: 2022 Citation: Chaulagain, B., Contina, J.B., Mills, K.B., Lantz, R.L., and Mundt, C.C. 202_. Comparing the efficacy of control strategies for infectious disease outbreaks using field and simulation studies. Ecol. Appl.:Under revision.
• Type: Journal Articles Status: Other Year Published: 2022 Citation: Contina, J.B., Lantz, R.L., Chaulagain, B., Mills, K.B., Tildesley, M.J., Keeling, M.J., and Mundt, C.C. 202_. The effect of farm connectedness and control strategies on the spread of foot-and-mouth disease of livestock. being reformatted for Ecosphere.
• Type: Journal Articles Status: Other Year Published: 2022 Citation: Seibel R.L., Meadows, M.J., Contina, J.B., Mundt, C.C., Keeling, M.J., and Tildesley, M.J. 202_. Modeling target-density-based cull strategies to contain foot-and-mouth disease outbreaks. Being reformatted for PeerJ.
• Type: Other Status: Other Year Published: 2022 Citation: Seibel R.L., Mills M.P., Nebert L., Mills K.B., and Mundt C.C. 202_.The effect of epidemic intensity on ring mitigation treatments for long-distance dispersed pathogens in a simulation study. In preparation.
• Type: Other Status: Other Year Published: 2022 Citation: Chaulagain, B., Contina, J.B., Mills, K.B., Lantz, R.L., and Mundt, C.C. 202_. Effect of host density on culling effects for control of wheat stripe rust in computer simulations. In preparation.
• Type: Other Status: Other Year Published: 2022 Citation: Mills, K.B., Seibel, R.L., and Mundt, C.C. 202_. Mechanisms contributing to the value of early interventions for disease control. In preparation.
• Type: Other Status: Other Year Published: 2022 Citation: Mundt, C.C., Chaulagain, B., Contina, J.B., Mills, K.B., Seibel, R.L., Moon, S.A., Cohnstaedt, L.W., Scoglio, C.M., Tildesley, M.J., Keeling, M.J., Jones, C.M., and Meentemeyer, R.K. 202_. Comparison of epidemic models for a common data field data set comparing mitigation methods. In preparation.
• Type: Other Status: Other Year Published: 2022 Citation: Mundt, C.C., Chaulagain, B., Contina, J.B., Mills, K.B., Seibel, R.L., Moon, Mill, M.P., Nebert, L., S.A., Cohnstaedt, L.W., Scoglio, C.M., Tildesley, M.J., Keeling, M.J., Jones, C.M., and Meentemeyer, R.K. 202_. Epidemic mitigation strategies modelled over a large, balanced parameter space. In preparation.
• Type: Other Status: Other Year Published: 2022 Citation: Jones, C.M., Chaulagain, B., Contina, J.B., Mills, K.B., Seibel, R.L., Moon, Mill, M.P., Nebert, L., S.A., Cohnstaedt, L.W., Scoglio, C.M., Tildesley, M.J., Keeling, M.J., Jones, C.M., and Meentemeyer, R.K., and Mundt, C.C. 202_. Epidemic mitigation strategies as influenced by host density, and host spatial pattern and pathogen fitness parameters. In preparation.

Progress 09/01/19 to 08/31/20

Outputs
Target Audience:1) Scientists, students, and stakeholders who have read scientific publications resulting from the project. 2) Policymakers, California Oak Mortality Task Force, and other stakeholders. 3) All scientists who are members of this projects, via inter-institutional visits and monthly Zoom electronic meetings. This includes a two day visit of Mike Tildsely (co-PI at University of Warwick) to Oregon State University, a four-day visit of Chris Jones (post-doc at NC State)) to Oregon State University, monthly Zoom meetings, and many informal Zoom meetings and face-face-meetings at each institution prior to COVD-19. 4) Stakeholders and scientists who attended Oregon Sudden Oak Death Task Force meeting in Marion County in September 2019. 5) Land managers and other scientists who used the tangible landcape model to study epidemic mitigations. 6) Scientists who heard the following presentation: Lantz, R.L., Meadows, A.J., Contina, J.B., Mundt, C.C., Keeling, and Tildesley, M.J. Doing more with less: Combating foot-and-mouth disease with target density-based culling. Salt Lake City, Utah, 2-7 August 2020 (accepted for oral presentation). 7) Students in classes taught by the PIs. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?1) Training in modeling and experimental design was provided to 3 postdocs, 2 graduate students, and 2 technicians. 2) Three postdocs, 2 graduate students, and 2 technicians received significant experience/training in writing of scientific publications. 3) One Ph.D student supported by the project completed his degree. 4) Chris Jones (Research Associate, North Carolina State University) traveled to Oregon State University Oregon and met with multiple project personnel for four days to incorporate other disease system specifications into the PoPS model.; 2 postdocs and 2 technicians from Oregon State University benefitted from this interaction. 5) Mike Tildesley (co-PI, University of Warwick) spent two days at Oregon State University in early February 2020 to discuss FMD and across systems modeling issues, before he had to return to the UK to address the unfolding COVID-19 epidemic; 2 postdocs and 2 technicians from Oregon State University benefitted from this interaction. 6) A postdoc and graduate student attended the Oregon Sudden Oak Death Task Force in Marion County in September 2019, we demonstrated two kinds of user interface fora new forecasting technology called PoPS--Pest or Pathogen Spread--that can help formulate management strategies. 7) We have held monthly group Zoom meetings for all project personnel throughout the reporting year, as well as many individual Zoom meetings with smaller groups to discuss specific project goals, data analyses, and modeling issues. 8) A postdoc attended Annual Meeting of the American Phytopathological Society (held virtually). 9) A technician attended the Annual Meeting of the Ecological Society (held virtually) and made a presentation (competitively chosen for oral presentation). How have the results been disseminated to communities of interest?1) Results were disseminated to scientists through publication of 8 scientific journal articles, 1 software package release, and 1 presentation at a scientific meeting. 2) During the project year, we have made significant strides in our participatory modeling work with Oregon stakeholders managing sudden oak death disease spread. At a gathering of the Oregon Sudden Oak Death Task Force in Marion County in September 2019, we demonstrated two kinds of user interface fora new forecasting technology called PoPS--Pest or Pathogen Spread--that can help formulate management strategies.This is a follow up to a workshop in 2017, the results of which are now published in Philosophical Transactions of the Royal Society B. Over thirty meeting attendees from government, academia, nonprofit groups and the timber industry experimented with the two interfaces and offered their feedback on PoPS, as they tested different "what-if" scenarios to curb disease spread. One interface usesTangible Landscapetechnology with a touchable, interactive surface to run PoPS on a local computer, while the other is a web platform that runs PoPS in the cloud. The web platform and Tangible Landscape both run the PoPS model, a sophisticated computer simulation that uses information about existing infections, locations of host trees and weather conditions to predict the spread of sudden oak death disease. The PoPS forecasting system allows users to interact with that simulation, and practice spending a realistic budget to try to prevent the spread. Both interfaces store those management choices in a cloud-based database and share a visualization dashboard for easily comparing the effectiveness of different strategies. 3) Results from the sudden oak death project have also been disseminated through blogs and newsletters. 4) Wheat producers were exposed to the goals of the project at virtual field days and informal discussions. 5) Outreach activities have been significantly limited during this reporting period owing to travel restrictions and other interference associated with the COVID-19 epidemic. What do you plan to do during the next reporting period to accomplish the goals?1) A major emphasis will be on completing the comparison of the effects of mitigation strategies on epidemic spread under a wide range of epidemiological parameters that incorporate all of the diseases being considered in this project. There will be a focus on evaluating the effects of specific biological systems from their component epidemiological parameters. Based on work of the past year, it now appears that we will be able to to utilize models from four different research groups in a factorial to compare spread and mitigation of foot-and- mouth disease of livestock, sudden oak death, West Nile Virus, and wheat stripe rust. 2) The above work will be used to develop heuristic rules of thumb and generalized theory for evaluating disease spread for the widest possible range of pathogens that spread via long distance dispersal. This will provide a framework for developing appropriate mitigation strategies under different environmental conditions and for different pathogen biology.

Impacts
What was accomplished under these goals? 1) A major, critical effort during the reporting period was in modifying models and choosing appropriate validation data sets to compare epidemic spread and epidemic mitigation across the diverse disease systems incorporated in this project. For example, we have needed to incorporate the ability to implement a diversity of control tactics in some of the models, and the ability to incorporate latent, infection periods, and other factors into others. To aid in this effort, a two-day Zoom workshop was held in July 2020 with all project personnel to work out the necessary details. 2) Field experiments of the previous year were repeated at two locations to determine effects of the timing and extensiveness of culling and population protection of the spread and control of wheat stripe rust. Results showed that cull timing was more important than cull size and, not surprisingly, mass fungicide application had the largest impact. 3) The field data described above was used to validate a spatially explicit, stochastic simulation model, which was then used to study a much wider range of mitigation tactics and epidemiological parameters/conditions than is possible in the field. These simulations showed that the relationships found in the field experiments hold over a much broader range of spatial and temporal variation in mitigation treatments. 4) Effects of host density on epidemic progression and efficacy of epidemic mitigations is under investigation. There is substantial variation in density of hosts and number of infection sites among diseases. For example, there are as many potential infection sites of wheat stripe rust in 20 hectares of wheat as there are people in the world. The current modeling project is showing that, not surprisingly, host density impacts epidemic progression. Much more importantly, the work is showing in spatially explicit, stochastic simulations that epidemic interventions are much more effective in low density host populations than in high density populations. This is of signal importance in understanding control strategies for epidemic of hosts of differing density. 5) Using wheat stripe rust as model system initially, we are uncovering the epidemiological mechanisms to explain the extreme importance of the timing of mitigation strategies on their success that have been uncovered by our work. For example, we have recently found through spatially explicit simulations that descendants of spores produced in the first three days of a wheat stripe rust epidemic account for about 85% of the total disease that occurs over the next four generations, and that this can be explained by very simple population growth models. We are currently determining how these relationships change with different epidemic rates and levels of initial disease prevalence, and how these factors impact disease control. We are currently in a second stage of this modeling work, in which we will study the spread and control of disease using epidemiological inputs relevant to COVID-19, with epidemic mitigations being implemented at different times after epidemic initiation, and when mitigations such as social distancing and contact tracing are utilized. 6) A network model is being used to study the spread and mitigation of COVID-19 in Manhattan, Kansas using the spatial distribution of actual household data. 7) One of our co-PIs has been asked to consult with a group that is attempting to study the spread of COVID-19 throughout all of Syracuse, NY, based on DNA and RNA samples collected from the sewer system, definitely to the level of clusters of homes, and possibly to the level of individual homes. We may be able to access these data for use in our models. 8) Vectorial capacity was modeled and assessed for risk of vector-borne disease spread using a spatiotemporal network model and climate data with an application of dengue in Bangladesh. 9) A generalizable model for forecasting the spread of plant diseases with interactive management was built. Additionally, we have tested these features with multiple user groups in order to ensure that the management interventions simulated are realistic. 10) A mean-field group-based epidemic spreading framework was developed to reduce the computational time of our individual based Generalized EpidemicModelingFramework (GEMF) framework. 11) GEMF network models were applied to Ebola and Africa swine fever transmission. 12) A network-based model (stochastic and deterministic) was developed for wheat stripe rust. 13) The beginnings of a project were initiated to compare four different spatiotemporal models to determine the degree to which they replicate effects of mitigation practices quantified in field experiments of wheat stripe rust.

Publications

• Type: Journal Articles Status: Published Year Published: 2020 Citation: Anna Petrasova, Devon A. Gaydos, Vaclav Petras, Chris M. Jones, Helena Mitasova, Ross K. Meentemeyer, Geospatial simulation steering for adaptive management, Environmental Modelling & Software, Volume 133, 2020, 10480.
• Type: Journal Articles Status: Accepted Year Published: 2021 Citation: Chris Jones, Devon Gaydos, Anna Petrasova, Vaclav Petras, Shannon Jones, Yu Takeuchi, Megan Skrip, and Ross Meentemeyer, Iteratively Forecasting Invasions with PoPS and a Little Help from Our Friends, Frontiers of Ecology and Environment
• Type: Journal Articles Status: Accepted Year Published: 2021 Citation: Moon SA, Sahneh FD, Scoglio C. Group-based general epidemic modeling for spreading processes on networks: GroupGEM. Accepted in the IEEE Transactions on Network Science and Engineering
• Type: Journal Articles Status: Under Review Year Published: 2021 Citation: Riad, MH; Cohnstaedt, L; Scoglio, CM. Modeling vectorial capacity and assessing risk for vector-borne disease spreading using spatiotemporal network model and climate data with an application of dengue in Bangladesh. Under second review in the American Journal of Tropical Medicine and Hygiene.
• Type: Journal Articles Status: Under Review Year Published: 2021 Citation: Severns, P.M., Sackett, K.E., and Mundt, C.C.. Single-focus outbreak suppression and the interactions between treatment timing and area for a long distance dispersed pathogen. Scientific Reports..
• Type: Books Status: Published Year Published: 2020 Citation: MMH Riad. 2020. Network-based Modeling for Risk Assessment of Infectious Disease Transmission. Ph.D. Thesis, Kansas State University.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Ferdousi T, Moon SA, Self A, Scoglio C (2019) Generation of swine movement network and analysis of efficient mitigation strategies for African swine fever virus. PLoS ONE 14(12): e0225785.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Riad, M.H., Sekamatte, M., Ocom, F. et al. Risk assessment of Ebola virus disease spreading in Uganda using a two-layer temporal network. Sci Rep 9, 16060 (2019).

Progress 09/01/18 to 08/31/19

Outputs
Target Audience:1) All scientists who are members of this projects, via interinstitutional visits and monthly Zoom electronic meetings.This includes a two day visit of Oregon State University to Kansas to learn about their models, 2) Stakeholders and scientists who heard the following presentations: a. Gaydos, D.A., Jones, C., Meentemeyer, R.K. 2019. Dynamic Geospatial Simulations for Assessing Landscape-Scale Control of Forest Disease Spread in Oregon. US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. b. Jones, C.J., Gaydos, D.A., Petras, V., Petrasova, A., and Meentemeyer, R.K., Moody, A. "Dynamic Species Distribution Modeling for Management of Emerging Pests and Pathogens." US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. c. Kruskamp, N. F., Gray, J.M., and Meentemeyer, R.K. 2019. Can Satellite Imagery Help Prioritize Forest Disease Management? US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. d. Gaydos, D.A.,Petrasova, A., Jones, C.M., Cobb, R.C.,Meentemeyer, R.K. The development and application of dynamic, geospatialP.ramorumspread models for Oregon. 7thSudden Oak Death Symposium, San Francisco, CA. e. Gaydos, D.A..The Social Side of Forest Disease Modeling. The Center for Geospatial Analytics Geospatial Forum, Raleigh, NC. f. Short-term forecast and dual state-parameter estimation for Japanese Encephalitis transmission using ensemble Kalman filter. In 2019 American Control Conference, July 10, Philadelphia, PA g. A spatio-temporal individual-based network framework for West Nile virus in the USA: Spreading pattern of West Nile virus, In iREDEFINE Workshop, ECEDHA Conference, AZ 3) Students who heard the following presentations: a. Chris Jones and Devon Gaydos helped guest lecture a class about disease modeling for a Grand Challenges of Geospatial Modeling class at NC State University lead by Dr. Ross Meentemeyer. b. Chris Jones gave an invited lecture to the Food Security and Emerging Diseases Class at NCSU lead by Dr. Jean Ristanio. c. Devon Gaydos also guest lectured for NCSU GIS 714 on the topic of calibration and validation of epidemiological simulations. 4) Scientists who have read the 11 refereed journal articles and 1 conference proceedings produced by this project during the reporting period. Changes/Problems:We are currently a bit behind schedule on project accomplishments owing to postdocs obtaining permanent positions before their terms were completed, and in replacement postdocs leaving for permanent positions or receiving a large grant as lead PI simultaneous with being offered postdocs on this project. We thus had signficant gaps of time without crucial postdoc help. We are know fully staffed with very capabale individuals, and have also hired a full-time research assistant to help with the modeling efforts. We will be able to fully complete the goals of the project through using a one-year no-cost extension. What opportunities for training and professional development has the project provided?1) Training in modeling and experimental design was provided to 3 postdocs, 3 graduate students, and 3 technicians. 2) A modeling exchange was held through a visit by Oregon personnel to Kansas State University. 3) Three postdocs, 3 graduate students, and 3 technician received siginficant experience/training in writing of scientific publications. 4) Postdocs and graduate students were involved in presenting six conference talks, three class lectures, and one other presentation 5) Experience was gained from monthly Zoom meetings among personnel from all four universities How have the results been disseminated to communities of interest?1) Results were disseminated through publication of 11 scientific journal articles, one conference procedding paper, and one software package release .2) Results were communicated through the following presentations: Gaydos, D.A., Jones, C., Meentemeyer, R.K. 2019. Dynamic Geospatial Simulations for Assessing Landscape-Scale Control of Forest Disease Spread in Oregon. US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. Jones, C.J., Gaydos, D.A., Petras, V., Petrasova, A., and Meentemeyer, R.K., Moody, A. "Dynamic Species Distribution Modeling for Management of Emerging Pests and Pathogens." US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. Kruskamp, N. F., Gray, J.M., and Meentemeyer, R.K. 2019. Can Satellite Imagery Help Prioritize Forest Disease Management? US-IALE Annual Meeting, April 7-11, 2019, Fort Collins, CO. Gaydos, D.A.,Petrasova, A., Jones, C.M., Cobb, R.C.,Meentemeyer, R.K. The development and application of dynamic, geospatialP.ramorumspread models for Oregon. 7thSudden Oak Death Symposium, San Francisco, CA. Gaydos, D.A..The Social Side of Forest Disease Modeling. The Center for Geospatial Analytics Geospatial Forum, Raleigh, NC. Short-term forecast and dual state-parameter estimation for Japanese Encephalitis transmission using ensemble Kalman filter. In 2019 American Control Conference, July 10, Philadelphia, PA A spatio-temporal individual-based network framework for West Nile virus in the USA: Spreading pattern of West Nile virus, In iREDEFINE Workshop, ECEDHA Conference, AZ 3) Results were communicated through the following class lectures: Chris Jones and Devon Gaydos also helped guest lecture a class about disease modeling for a Grand Challenges of Geospatial Modeling class at NC State University lead by Dr. Ross Meentemeyer. In addition, Chris Jones gave an invited lecture to the Food Security and Emerging Diseases Class at NCSU lead by Dr. Jean Ristanio. Devon Gaydos also guest lectured for NCSU GIS 714 on the topic of calibration and validation of epidemiological simulations. 4) Wheat producers were exposed to the goals of the project. 5) Scientists and land managers learned of our work through use of the the Tangible Landscape modeling tool. What do you plan to do during the next reporting period to accomplish the goals?1) A major emphasis will be on comparing the effects of mitigation strategies on epidemic spread under a wide range of epidemiological parameters that incorporate all of the diseases being considered in this project.. There will be a focus on evaluating the effects of specific biological syytems from their component epidemiological parameters so as to develop heuristic rules of thumb for evaluating the widest possible range of pathogens that spread via long distance dispersal. At the moment, the PoPS model seems like the best platform for modeling all diseases simultaneously, but comparisons also will be made with combinations of two or three models for specific disease systems. 2) Field plots were establsihed at two locations to study effects of cull timing, cull size, crop protection with fungicide, and host dneisty on epidemic spread. 3) Data will be collected from the wheat stripe rust field experiments. 4) The models EPIMUL and PoPS will be used to further expand upon results obtained from the 5) We have developed a generalized group based framework to understand the diseases spreading on very large networks (millions of nodes), called GgroupEM. First, we will evaluate the performance of this framework on different scenarios (different sizes and types of networks, different initial conditions, different disease parameters, etc.). Second, we will develop a Monte Carlo simulation-based tool for the exact model of the group-based framework. This tool, called GgroupEMsim, will be similar to GEMFsim but for extremely large networks. 6) We will develop a risk assessment framework for vector-borne diseases. In this framework, we will explicitly use weather data such temperature and rainfall to model the mosquito vectorial capacity, which is very crucial in the risk assessment process. We will develop a spatially explicit network model and simulate spatio-temporal spreading of vector-borne diseases. 7) We are nearing the release of Version 1.0 of the PoPS Forecasting system and are analyzing the data from the workshop for both a paper and as part of an economic analysis for the SOD Task Force to present to state congress to request a budget increase to manage SOD based on some of the findings from the workshop. 8) Work will be completed on studies to explore the possibility of refining the scorched-earth policy with foot-and-mouth disease to uncover alternative cull-based control strategies that provide equal protection to the scorched-earth policy without as many control causalities.

Impacts
What was accomplished under these goals? 1) We developed an interface that usesTangible Landscapetechnology with a touchable, interactive surface to run PoPS on a local computer, while the other is a web platform that runs PoPS in the cloud. The web platform and Tangible Landscape both run the PoPS model, a sophisticated computer simulation that uses information about existing infections, locations of host trees and weather conditions to predict the spread of sudden oak death disease. The PoPS forecasting system allows users to interact with that simulation, and practice spending a realistic budget to try to prevent the spread. Both interfaces store those management choices in a cloud-based database and share a visualization dashboard for easily comparing the effectiveness of different strategies. 2) We are nearing the release of Version 1.0 of the PoPS Forecasting system and are analyzing the data from the workshop for both a paper and as part of an economic analysis for the SOD Task Force to present to state congress to request a budget increase to manage SOD based on some of the findings from the workshop. 3) We have estimated the swine movement network at farm level in the US. We found that high in-degree farm operations (i.e., markets) play critical roles in the disease spread. We also found that high in-degree based targeted isolation and hypothetical vaccinations are more effective for disease control compared to other centrality-based mitigation strategies. 4) We developed a generalized group-based epidemic model (GgroupEM) framework for any compartmental epidemic model (for example; susceptible-infected-susceptible, susceptible-infected-recovered, susceptible-exposed-infected-recovered).Since the group-based framework is computationally less expensive and faster than an individual-based framework, then this framework is useful when the simulation time is important. 5) Developed a risk assessment framework for Ebola using a two-layer temporal network.Simulation results show that some regions are at higher risk of infection, suggesting some focal points for Ebola preparedness and providing direction to inform interventions in the field. Simulation results also show that decreasing physical contact as well as increasing preventive measures result in a reduction of chances to develop an outbreak. Overall, the main contribution of this paper lies in the novel method for risk assessment, which can be more precise with an increasing volume of accurate data for creating the network model. 6) Developed an Individual-level network model for RVF in Uganda.. Our goal was to investigate changes in the epidemic size (total number of infected cows) for varying mosquito abundance, different initial conditions (single or multiple outbreak locations), cattle breed (indigenous or exotic), and cattle movement. We suggest several mitigation strategies to check/reduce RVFV spread using simulation results from the individual-based network model. 7) We formulated an ensemble Kalman filter (EnKF) that provides dual state-parameter estimates for the transmission of Japanese Encephalitis (JE) - a vector-borne disease.We found that the EnKF method for forecasting vector-borne disease spread should only be used for two to four time steps, i.e. real-time forecast during an outbreak when new incidence data available continuously. We also demonstrated the effectiveness of control measures on the epidemic by simulating various vector population abundance. 8) Completed field experiments at two locations to compare effects of cull timing, cull size, and fungicide protection on the spread of wheat stripe rust. Preliminary analyses suggest that protection is the most effective mitgation strategy and that cull timing is more important than cull size. 9) Parametrized the spatially explicit simulation model EPIMUL, which repreoduced well results from the wheat stripe rust field experiments. This will give us confidence in modeling variables and ranges of variables not possible to include in the field experiments. 10) We continued with studies to explore the possibility of refining the scorched-earth policy with foot-and-mouth disease to uncover alternative cull-based control strategies that provide equal protection to the scorched-earth policy without as many control causalities. 11) Developed a draft data set of epidemic inputs for the purpose of modeling effects of epidemic parameters and mitigation strategies on epidemic spread. 12) Began comparison of two different models (from Kansas and UK) on spread and control of foot-and-mouth disease in geographic areas with different spatial patterns of livestock farms. 13) Significant effort was invested in beginning the process of identifying appropriate models for studying epidemic parameters and mitigation strategies on epidemic spread, with the goal of identifying generalized rules of thumb across LDD pathogens for the most effective approaches to epidemic mitigation for a given set of parameters.

Publications

• Type: Journal Articles Status: Under Review Year Published: 2019 Citation: Sifat Afroj Moon, Faryad Darabi Sahneh, and Caterina Scoglio. Generalized group-based epidemic model for spreading processes on networks: Ggroupem. arXiv preprint arXiv:1908.06057, 2019. (Under review on IEEE Transaction on Network Science and Engineering)
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Sifat A Moon, Tanvir Ferdousi, Adrian Self, and Caterina M Scoglio. Estimation of swine movement network at farm level in the USA from the census of agriculture data. Scientific reports, 9(1):6237, 2019.
• Type: Journal Articles Status: Accepted Year Published: 2019 Citation: Tanvir Ferdousi, Sifat Afroj Moon, Adrian Self, and Caterina Scoglio. Generation of swine movement network and analysis of efficient mitigation strategies for African swine fever virus. arXiv preprint arXiv:1911.04447, 2019. (Accepted in PlosOne)
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Musa Sekamatte, Mahbubul H Riad, Tesfaalem Tekleghiorghis, Kenneth J Linthicum, Seth C Britch, Juergen A Richt, JP Gonzalez, and Caterina M Scoglio. Individual-based network model for rift valley fever in Kabale district, Uganda. PloS one, 14(3):e0202721, 2019.
• Type: Conference Papers and Presentations Status: Published Year Published: 2019 Citation: 5Mahbubul H Riad, Caterina M Scoglio, Lee W Cohnstaedt, and D Scott McVey. Short-term forecast and dual state-parameter estimation for Japanese encephalitis transmission using ensemble Kalman filter. In 2019 American Control Conference (ACC), pages 34443449. IEEE, 2019.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Mahbubul H Riad, Musa Sekamatte, Felix Ocom, Issa Makumbi, and Caterina M Scoglio. Risk assessment of Ebola virus disease spreading in Uganda using a two-layer temporal network. Scientific Reports (Nature Publisher Group), 9:117, 2019.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Sifat A Moon, Lee W Cohnstaedt, D Scott McVey, and Caterina M Scoglio. A spatio-temporal individual-based network framework for west Nile virus in the USA: Spreading pattern of west Nile virus. PLoS computational biology, 15(3):e1006875, 2019.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Tanvir Ferdousi, Lee W Cohnstaedt, D Scott McVey, and Caterina M Scoglio. Understanding the survival of Zika virus in a vector interconnected sexual contact network. Scientific reports, 9(1):7253, 2019.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Forecasting and Control of Emerging Infectious Disease through Participatory Modelling. D. Gaydos, A. Petrasova, R. Cobb, R. Meentemeyer. Philosophical Transactions of the Royal Society B. May 2019.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Direct and indirect effects of forest microclimate on pathogen spillover. Dillon, W. and Meentemeyer, R.K. Ecology. March 2019
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Severns, P.M., Sackett, K.E., Farber, D.H., and Mundt, C.C. 2019. Consequences of long-distance dispersal for epidemic spread: Patterns, scaling, and mitigation. Plant Dis. 103:177-191 (invited Feature Article).
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Farber, D.H., De Leenheer, P., and Mundt, C.C. 2019. Dispersal kernels may be scalable: Implications from a plant pathogen. J. Biogeogr. 46:2042-2055.

Progress 09/01/17 to 08/31/18

Outputs
Target Audience:Target audiences for this reporting period include: 1) Scientists who are particpants in this study, via investigator meeting in Corvallis, Oregon, May2018. 2) Faculty, postdocs, students, and other scientistswho heard the following seminars: Moon, S.A. Understanding of the spreading pattern of West Nile virus in the USA using approximate Bayesian computation, November 2018, Electrical and Computer Engineering Deptartment, Kansas State University. Severns, P.M.Department of Plant Pathology, University of Georgia, 2018, invited seminar: "Long-Distance Dispersal and Disease Spread" 3)Faculty, postdocs, students, and other scientists who heard the following presentations: Gaydos, D.A., Petrasova, A., Petras, V., Cobb, - -R.C., Meentemeyer, R.K. 2018. Managing forest disease spread with tangible participatory models. 9th International Congress on Environmental Modelling and Software, Fort Collins, CO Gaydos, D.A., Petrasova, A., Petras, V., Cobb, R.C., Meentemeyer, R.K. 2018. Managing forest disease spread with Tangible Landscape technology. US-IALE Annual Meeting, Chicago, IL. Ferdousi, T., Cohnstaedt, L., Mcvey, D., and Scoglio, S. Understanding the role of sexual transmission in the spread of ZIKA virus using an individual-based interconnected population model, February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City. Gaydos, D.A. and Meentemeyer, R.K. 2017. Tangible geospatial models as decision support systems for forest disease mitigation. Southern Forestry GIS Conference, Athens, GA. Jones, C.J., Tonini, F., Meentemeyer, R.K., Moody, A. "Interactions and impacts of sudden oak death and fire using coupled dynamics spatial temporal epidemiological modeling in Big Sur, CA." US-IALE Annual Meeting, April 8-11, 2018, Chicago, IL. Moon, S.F., Cohnstaedt, L., and Scoglio, C.M. A spatio-temporal individual-based network framework for West Nile virus in the USA: spreading pattern of West Nile virus", February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City Riad, M., Cohnstaedt, L., Mcvey, D., and Scoglio, S. Forecasting infectious disease outbreak using filtering framework, February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City. 4) Studensts who heard aguest lecture a class about disease modeling for a Grand Challenges of Geospatial Modeling class at NC State University. 5) Scientists who read scientific papers produced by this project. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?1) Training in modeling and experimental design was provided to 3 postdocs, 3graduate students, and 1 technician. 2) The investigator meeting provided an opportunity for all senior personnel, 3 postdocs, 3graduate students, and 1 technician to interact in a multidisciplinary gathering. 3) Three postdocs, 3graduate students, and 1 technician received siginficant experience/training in writing of scientific publications. 4) Experience was gained from several scientific presentations resulting from the work. 5) One postdoc received experience in international collaboration through interactions with the UK collaborators. How have the results been disseminated to communities of interest?1) An investigator meeting was held in in Corvallis, Oregon, May 2018. 2) Results were communicated through the following seminars: Moon, S.A. Understanding of the spreading pattern of West Nile virus in the USA using approximate Bayesian computation, November 2018, Electrical and Computer Engineering Deptartment, Kansas State University. Severns, P.M.Department of Plant Pathology, University of Georgia, 2018, invited seminar: "Long-Distance Dispersal and Disease Spread" 3) Results were communicated through the following presnetations:: Gaydos, D.A., Petrasova, A., Petras, V., Cobb, - -R.C., Meentemeyer, R.K. 2018. Managing forest disease spread with tangible participatory models. 9th International Congress on Environmental Modelling and Software, Fort Collins, CO Gaydos, D.A., Petrasova, A., Petras, V., Cobb, R.C., Meentemeyer, R.K. 2018. Managing forest disease spread with Tangible Landscape technology. US-IALE Annual Meeting, Chicago, IL. Ferdousi, T., Cohnstaedt, L., Mcvey, D., and Scoglio, S. Understanding the role of sexual transmission in the spread of ZIKA virus using an individual-based interconnected population model, February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City. Gaydos, D.A. and Meentemeyer, R.K. 2017. Tangible geospatial models as decision support systems for forest disease mitigation. Southern Forestry GIS Conference, Athens, GA. Jones, C.J., Tonini, F., Meentemeyer, R.K., Moody, A. "Interactions and impacts of sudden oak death and fire using coupled dynamics spatial temporal epidemiological modeling in Big Sur, CA." US-IALE Annual Meeting, April 8-11, 2018, Chicago, IL. Moon, S.F., Cohnstaedt, L., and Scoglio, C.M. A spatio-temporal individual-based network framework for West Nile virus in the USA: spreading pattern of West Nile virus", February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City Riad, M., Cohnstaedt, L., Mcvey, D., and Scoglio, S. Forecasting infectious disease outbreak using filtering framework, February 2018, AMCA (American Mosquito Control Association) 85th annual meeting, Kansas City. 4) Results were communicated to scientists through peer-reviewed publications. 5) Graduate students were exposed to the work in the Oregon State University course entitled "Plant Disease Management", Winter 2018. 6) Wheat producers were exposed to the goals of the project. 7) Scientists and land managers learned of our work in a workshop with key local stakeholders in October 2017 where we demonstrated the Tangible Landscape modeling tool. 8) Our work was presented to stakeholders at Oregon State University, Oregon Department of Forestry, and US Forest Service as part of an informal Sudden Oak Death Symposium. 9) Work on disease modeling was presneted in a Grand Challenges of Geospatial Modeling class at NC State University. What do you plan to do during the next reporting period to accomplish the goals?1) Modeling work will continue regarding effects of initial disease prevalence and basic reproduction number of long-distance disease spread. 2) Studies of designed experiments and natural experiments will be used to study effects of initial disease prevalence and basic reproduction number of long-distance disease spread. 3) Field experiments with wheat stripe rust will be conducted to determine the relative efficacy of cull timing, cull size, and extent of population protection in suppressing epidemic outbreaks. 4) Modeling efforts will be initiated with wheat stripe rust to determine the effect of host density on the efficacy of epidemic interventions. 5) We will develop a swine movement network for the US swine industry. In this network, we will apply the different culling processes and analyze the outbreak size. 6) Continue use of tangible landscape model and other platforms developed at NC State to study commonalities and differences among disease sysytems. 7) Interactions of initial disease prevalence, basic reproduction number, and culling strategies will be studied in mathematical studies of all diseases. 8) Studies will be completed to compare pathogen dispersal kernels among the diseases being studied; this will be presented in an invited article in Annual Review of Phytopathology. 9) issues of spatial scaling of long-distance disease spread will be addressed mathematically for all diseases.

Impacts
What was accomplished under these goals? The impact of this project is to provide improved information to better manage outbreaks of disease caused by pathogens that are capable of long-distance movement. The ultimate goal is to provide generalizations or "rules-of-thumb" that will enable epidemiologists and health care officials to rapidly predict patterns and rate of spread, and the efficacy of different outbreak interventions. This information will be very helpful in more effectively making the fast decsisions that are necessary during an epidemic outbreak. The commonalities about different diseases that we are studying may help to identify disease control approaches that are similar among plant, animal, and human diseases. Progress this reporting period: 1) Field studies with wheat stripe rust were conducted to determine effects of initial disease prevalence and basic infection number on the efficacy of culling for epidemic suppression. We found that culling had similar effects regardless of initial disease prevalence and basic infection number. This is consistent with simulation studies that were done in the previous reporting period. 2) Field studies were established to determine effects of cull timing, cull size, and protection in different spatial arrangements on the spread of wheat stripe rust. Plots will be inoculated and treatments applied in spring. 3) We used a previously developed spatial FMD model (Keeling et al. 2001) to simulate hypothetical FMD outbreaks and control strategies in a region with farm demography similar to Cumbira, UK, which was one of the epidemic hotspots during the 2001 epidemic. Our aim was to explore the possibility of refining the scorched-earth policy to uncover alternative cull-based control strategies that provide equal protection to the scorched-earth policy without as many control causalities. Although we uncovered small reductions in the number of farms culled in the target density strategy compared to the scorched-earth strategy, we think our results demonstrate the potential for more refined and targeted cull strategies to lessen the impact of control measures without compromising control of the epidemic. 4) We have proposed an individual based heterogeneous network model to understand the spreading pattern of the WNV in the USA. We found that a modified fat-tailed dispersal kernel could describe the human incidence data better than general fat-tailed kernel network model. We proposed some control measures for this disease; our simulation suggests that mosquito population reduction in the infected states with its neighboring states is potentially cost-effective. 5) We proposed a novel individual-based interconnected network model that incorporates both insect-vectored and sexual transmission of the Zika virus. We studied the outbreak size with different initial conditions. Our results suggest that although sexual transmission has a relatively low contribution in determining the epidemic size, it plays a role in sustaining the epidemic and creating potential endemic scenarios. 6) We evaluated three filtering approaches to estimate Japenese encephalitis transmission rate and for a forecasting of the disease outbreak. First, we used the ensemble Kalman filter (EnKF) to make a short-term forecast using a constant parameter set. Second, we used a dual state-parameter estimation framework based on ensemble Kalman filter to simultaneously estimate the parameters and states (no of individual in compartment). Finally, we proposed the CPEFF, in which we integrated the concept of kernel density particle filter (KDPF) and ensemble Kalman filter (EnKF). 7) During this fiscal year, we have made significant strides in our participatory modeling work with Oregon stakeholders managing sudden oak death disease spread. We have since analyzed all survey data collected at the first workshop, and have made substantial changes to the epidemiological model, including adding mortality and acquiring finer resolution host and weather data.

Publications

• Type: Journal Articles Status: Published Year Published: 2018 Citation: Meadows, A.J., Mundt, C.C., Keeling, M.J., and Tildesley, A.J. 2018. Disentangling the influence of livestock versus farm density on livestock disease epidemics. Ecosphere 9(7):e02294.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Mundt, C.C. 2018. Pyramiding for resistance durability: Theory and practice. Phytopathology (invited review) Phytopathology 108:792-802.
• Type: Journal Articles Status: Accepted Year Published: 2018 Citation: Severns, P.M., Sackett, K.E., Farber, D.H., and Mundt, C.C. 2018. Consequences of long-distance dispersal for epidemic spread: Patterns, scaling, and mitigation. Plant Disease (invited Feature Article):https://doi.org/10.1094/PDIS-03-18-0505-FE.
• Type: Journal Articles Status: Other Year Published: 2019 Citation: Farber, D.H., De Leenheer, P., and Mundt, C.C. 2019. Modeling the spread of a plant disease at varying spatial scales. Journal of Biogeography:In revision.
• Type: Journal Articles Status: Under Review Year Published: 2019 Citation: Gaydos, D., Petrasova, A., Cobb, R., and Meentemeyer, R. 2019. Forecasting and control of emerging infectious disease through participatory modelling. Philosophical Transactions of the Royal Society B.:Under review.
• Type: Journal Articles Status: Under Review Year Published: 2019 Citation: Moon, S.F., Cohnstaedt, L., and Scoglio, C.M. 2019. A spatio-temporal individual-based network framework for West Nile virus in the USA: spreading pattern of West Nile virus. PLoS Computational Biology:Under review.
• Type: Conference Papers and Presentations Status: Under Review Year Published: 2019 Citation: Riad, M., Cohnstaedt, L., Mcvey, D., and Scoglio, S. 2019. Short-term forecast and dual state-parameter estimation for Japanese Encephalitis transmission using ensemble Kalman filter. American Control Conference (ACC):Under review.
• Type: Journal Articles Status: Submitted Year Published: 2019 Citation: Ferdousi, T., Cohnstaedt, L., Mcvey, D., and Scoglio, S. 2019. Understanding the survival of Zika virus in a vector interconnected sexual contact network. Scientific Reports:Submitted.
• Type: Journal Articles Status: Submitted Year Published: 2019 Citation: Severns, P.M., Sackett, K.E., and Mundt, C.C. 2019. Single-focus outbreak suppression and the interactions between treatment timing and area for a long distance dispersed pathogen. PLoS Pathogens:Submitted.

Progress 09/01/16 to 08/31/17

Outputs
Target Audience:Target audiences for this reporting period include: 1) Scientists who are particpants in this study, via investigator meeting in Corvallis, Oregon, August 2017. 2) Faculty, posdocs, and students who heard the following seminars: Mundt, C.C. Department of Plant Pathology, University of Florida, 2017, invited seminar: "Long-Distance Dispersal and the Spread of Plant and Animal Diseases" Mundt, C.C. Plant Pathology Into the 21st Century (symposium in honor of Bill Fry), Cornell University, 2017. Invited speaker, short talk: "What I do on my summer vacations" 3) Scientists and land managers who heard the following presentations: Allison Simler, Margaret R. Metz, Ross K. Meentemeyer, Kerri M. Frangioso, Tyler. B. Bourret, and David M. Rizzo. 2017. Disturbance legacies and the post-fire dynamics of an emerging, introduced forest disease. Presnted at the annual meeting of the ecological society of America, August 2017. Meadows, A.J.,Mundt, C.C., Keeling, M.J., and Tildesley, M.J.. Annual meeting of the Ecological Society of America, Portland, OR 2017. Contributed oral presentation: "The impacts of farm and cattle density on foot-and-mouth disease epidemic spread". MH Riad, CM Scoglio, DS McVey, LW Cohnstaedt.Estimation of parameters and basic reproductive ratio for Japanese encephalitis transmission in the Philippines using a sequential Monte Carlo filter. Control Technology and Applications (CCTA), 2017 IEEE Conference on, 668-673. 4) Scientists who have read publications resulting from the work. 5) Graduate students who were exposed to the work in the Oregon State University course entitled "Plant Disease Dynamics", winter 2017. 6) Wheat producers who were exposed to the goals of the project. 7) Scientists and land managers who attended the Tangible Landscape Sudden Oak Death Management Workshop hosted by researchers at North Carolina State University and University of California, Davis and conducted in Salem, OR on October 10, 2017. A new geospatial tool was used to evaluate how sudden oak death management actions affect disease spread and impacts across the landscape. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?1) Training in modeling and experimental design was provided to 3 postdocs, 2 graduate students, and 1 technician. 2) The investigator meeting provided an opportunity for all senior personnel, 3 postdocs, 2 graduate students, and 1 technician to interact in a multidisciplinary gathering. 3) Three postdocs, 2 graduate students, and 1 technician received siginficant experience/training in writing of scientific publications. 4) Experience was gained from several scientific presentations resulting from the work. 5) One postdoc received experience in international collaboration through interactions with the UK collaborators. 6) One postdoc began preparation for management of a symposium on Disease Ecology to be held at Oregon State University. How have the results been disseminated to communities of interest?1) An investigator meeting was held in in Corvallis, Oregon, August 2017. 2) Results were communicated through the following seminars:: Mundt, C.C. Department of Plant Pathology, University of Florida, 2017, invited seminar: "Long-Distance Dispersal and the Spread of Plant and Animal Diseases" Mundt, C.C. Plant Pathology Into the 21st Century (symposium in honor of Bill Fry), Cornell University, 2017. Invited speaker, short talk: "What I do on my summer vacations" 3) Results were communicated to scientists and land managers inthe following presentations: Allison Simler, Margaret R. Metz, Ross K. Meentemeyer, Kerri M. Frangioso, Tyler. B. Bourret, and David M. Rizzo. 2017. Disturbance legacies and the post-fire dynamics of an emerging, introduced forest disease. Presnted at the annual meeting of the ecological society of America, August 2017. Meadows, A.J.,Mundt, C.C., Keeling, M.J., and Tildesley, M.J.. Annual meeting of the Ecological Society of America, Portland, OR 2017. Contributed oral presentation: "The impacts of farm and cattle density on foot-and-mouth disease epidemic spread". MH Riad, CM Scoglio, DS McVey, LW Cohnstaedt.Estimation of parameters and basic reproductive ratio for Japanese encephalitis transmission in the Philippines using a sequential Monte Carlo filter. Control Technology and Applications(CCTA), 2017 IEEE Conference on, 668-673. 4) Results were communicated to scientists through peer-reviewed publications. 5) Graduate students who were exposed to the work in the Oregon State University course entitled "Plant Disease Dynamics", winter 2017. 6) Wheat producers who were exposed to the goals of the project. 7) Scientists and land managers learned of our work in the Tangible Landscape Sudden Oak Death Management Workshop hosted by researchers at North Carolina State University and University of California, Davis and conducted in Salem, OR on October 10, 2017. A new geospatial tool was used to evaluate how sudden oak death management actions affect disease spread and impacts across the landscape. What do you plan to do during the next reporting period to accomplish the goals?1) Modeling work will continue regarding effects of initial disease prevalence and basic reproduction number of long-distance disease spread. 2) Studies of designed experiments and natural experiments will be used to study effects of initial disease prevalence and basic reproduction number of long-distance disease spread. 3) Effects of initial disease prevalence and basic reproduction number will be continued in designed experiments of wheat stripe rust and in natural experiments of foot-and-mouth disease, sudden oak death and arborviruses. 4) Interactions of initial disease prevalence, basic reproduction number, and culling strategies will be studied in mathematical studies of all diseases. 5) Field experiments with wheat stripe rust will be conducted to determine how initial disease prevalance and the basic reproduction number influence culling implemented for disease control. 6) Studies will be completed to compare pathogen dispersal kernels among the diseases being studied. 7) issues of spatial scaling of long-distance disease spread will be addressed mathematically for all diseases.

Impacts
What was accomplished under these goals? 1) A personnel meeting was held in Corvalls (and electronically) for personnel associated with the overall project. Many useful ideas were exchanged and research plans developed. 2) Large field studies were established to determine effects of initial disease prevalence, spatial pattern of initial disease prevalence, and basic infection number on spread of wheat stripe rust. 3) Models based on wheat stripe rust showed that initiall disease prevalence at the outbreak focus may play a larger role in subsquent disease spread than was originally anticipated by many. 4) Developed tangible landscape model to model SOD spread and compare control strategies. Shows real time epidemic outcomes, allows users to pause and rewind simulations, and compare control costs and efficacy in real time (i.e. users can adaptively move through the simulation). Can be adapted to other disease systems. 5) Developed a spatial metapopulation network model for West Nile Virus. Estimated network links based on local and migratory bird movement (network links are directed with respect to migration). Used reaction-diffusion process to model transmission between subpopulations, Beta is seasonal (depends on space and time). 6) Developed a state-space model to model dynamic data. This model has two parts: (1) observation equation (distribution of the number of cases at the current time step) and (2) state evolution equation (distribution of states at the current time step). 7) Developed a combined network model with human and mosquito transmitted Zika virus. Vector-coupled model. Model tries to understand the role of sexual transmission in Zika transmission. It is a simple model that has few parameters. It isis individual based and therefore has sexual transmission and vector-borne transmission simultaneously. 8) Stochastic models were used to study the spread of foot-and-mouth (FMD) disease using actual farm number and farm sizes for the states of Oregon, Pennsylvania, and Texas as well as more balance, theoretical data sets. In these simulations, increasing cattle density always resulted in larger epidemics, regardless of farm density. Alternatively, increasing farm density only led to larger epidemics in scenarios of high cattle density. 9) Preliminary modeling studies of FMD have been conducted to develop a strategy to minimize livestock losses due to epidemic control measures while maintaining control strategy efficacy. To do so, we compare the epidemic impact (livestock culled because of FMD infection + livestock preemptively culled due to control methods) and duration of the following control strategies: I. Density threshold strategy: 1.Caulculate the livestock density within the control radius of an infected farm. If it is below the density threshold, no culling is implemented. If not, proceed to step 2. 2.Sort farms from the largest to smallest in terms of the number of livestock. 3.Starting with the largest farm, go down the list and cull farms until the susceptible livestock density in the radius is below the threshold density. II. Universal strategy: Cull all the farms within the control radius of the infected farm. III. Combination strategy. Preliminary results suggesity that density threshhold strategy is not effective if farms are randomly distribuated in space, but can be highly effective when modeling actuall, aggregated spatial distributions of farms. 10) Continuing work on dispersal kerenels of pathogens that exhibit long distance dispersal is showing a high degress of similairity even among taxonomically divergent pathogens and hosts.

Publications

• Type: Journal Articles Status: Published Year Published: 2017 Citation: Farber, D.H., and C.C. Mundt. 2017. Effect of plant age and leaf position on susceptibility to wheat stripe rust. Phytopathology 107:412-417.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Farber, D.H., Medlock, J., and Mundt, C. C. 2017. Local dispersal of Puccinia striiformis f. sp. tritici from isolated source lesions. Plant Pathol. 66:28-37.
• Type: Journal Articles Status: Other Year Published: 2017 Citation: Meadows, A.J., Mundt, C.C., Keeling, M.J., and Tildesley, A.J. 201_. Disentangling the influence of livestock versus farm density on livestock disease epidemics. Ecosphere:accepted pending revisions.
• Type: Journal Articles Status: Other Year Published: 2017 Citation: Farber, D.H., De Leenheer, P., and Mundt, C.C. 201_. Modeling the spread of a plant disease at varying spatial scales. Ecol. Lett.:submitted.
• Type: Journal Articles Status: Other Year Published: 2017 Citation: Severns, P.M., Sackett, K.E., and Mundt, C.C. 201_. Culling impacts on spread of wheat stripe rust: Effects of timing and size of ring culls on spread of wheat stripe rust. PLoS Pathog.:Resubmission near completion.
• Type: Journal Articles Status: Other Year Published: 2017 Citation: Severns, P.M., Sackett, K.E., Farber, D.H., and Mundt, C.C. 201_. Consequences of long-distance dispersal for epidemic spread: Patterns, scaling, and mitigation. Plant Dis. (invited Feature Article):in preparation.

Progress 09/01/15 to 08/31/16

Outputs
Target Audience:Target audiences for this reporting period include: 1) Scientists who are particpants in this study, via an all-investigator meeting in Corvallis, Oregon, May 2016. 2) Scientists who heard a symposium presentation on the work at the annual meeting of the American Phytopathological Society in Tampa, FL, July 2016. 3) Scientists and land managers who heard the following presentations: Dillon, W.W., and Meentemyer, R.K. 2016. Influence of climate variability on pathogen spillover in a multi-host forest disease. US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Gaydos, D.A., and Meentemeyer, R.K. 2016.Resilience of the diversity-disease risk hypothesis following wildfire disturbance.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Jones, C.M., Moody, A., and Meentemeyer, R.K. 2016. Fire and disease interactions in the Sudden Oak Death system in Big Sur, California.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Tonini, F.,Shoemaker, D.,Harmon, B., and Meentemeyer, R.K. 2016. Collaborative solutions to invasive species management using Tangible Landscape.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. 4) Scientists who have read publications resulting from the work. 5) Graduate students who were exposed to the work in the Oregon State University course entitled"Plant Disease Management", winter 2016. 6) Wheat producers who were exposed to the goals of the project. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?1) Training inmodeling and experimental design was provided to 3 postdocs, 2 graduate students, and 1 technician. 2) The all-investigator meeting provided an opportunity for all senior peronnel, 3 postdocs, 2 graduate students, and 1 technician to interact in a multidisciplinary gathering. 3) Three postdocs, 2 graduate students, and 1 technician received siginficant experience/training in writing of scientific publications. 4) Experience was gained from several scientific presentations resulting from the work. 5) One postdoc received experience in international collaboration through travel and electronic communication with the UK collaborators. 6) One postdoc began preparation for management of a symposium on Disease Ecology to be held at Oregon State University. How have the results been disseminated to communities of interest?1) Scientists who are particpants in this study inmteractedinan all-investigator meeting in Corvallis, Oregon, May 2016. 2) A symposium presnetation entitled "Initial Epidemic Conditions and Long-Distance Spread of Disease" was presneted at the annual meeting of theof the American Phytopathological Society in Tampa, FL, July 2016. 3) Information was provided to scientists and land managers inthe following presentations: Dillon, W.W., and Meentemyer, R.K. 2016. Influence of climate variability on pathogen spillover in a multi-host forest disease. US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Gaydos, D.A., and Meentemeyer, R.K. 2016.Resilience of the diversity-disease risk hypothesis following wildfire disturbance.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Jones, C.M., Moody, A., and Meentemeyer, R.K. 2016. Fire and disease interactions in the Sudden Oak Death system in Big Sur, California.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. Tonini, F.,Shoemaker, D.,Harmon, B., and Meentemeyer, R.K. 2016. Collaborative solutions to invasive species management using Tangible Landscape.US International Association of Landsape Ecology Annual Meeting, Asheville, NC, April 3-7, 2016. 4) Scientists were exposed to journal article that have been published or submittedwho have read publications resulting from the work. 5) Graduate students were exposed to the work in the Oregon State University course entitled"Plant Disease Management", winter 2016. 6) Wheat producerswere exposed to the goals of the project during Extension presentations and informal interactions. What do you plan to do during the next reporting period to accomplish the goals?1) Modeling work will continue regardingeffects of initial disease prevalence and basic reproduction number of long-distance disease spread. 2) Studies of desgined experiments and natural experiments will be used to studyeffects of initial disease prevalence and basic reproduction number of long-distance disease spread. 3) Effects of initial disease prevalence andbasic reproduction number will be continued in designed experiments of wheat stripe rust andin natural experiments of foot-and-mouth disease, sudden oak death and arborviruses. 4) Interactions of initial disease prevalence, basic reproduction number, and culling strategies will be studied in mathematical studies of all diseases. 5) issues of spatial scaling of long-distance disease spread will be addressed mathematically for all diseases.

Impacts
What was accomplished under these goals? 1) An important first-year accomplishment for this large, multi-disciplinary project was simply in getting all positions staffed and in establishing multi-institutioncommunications. We had a very successful all-personnel meeting in Corvallis, OR in May 2016, where many useful ideas were exchanged and research plans developed. 2) Large field studies were establishedto determine effects of initial disease prevalence, spatial pattern of initial diseaseprevalence, and basic infection number on spread of wheat stripe rust. 3) Models based on wheat stripe rust showed that ring cull size has no impact on subsequent epidemic spread, while time of initiation of the cull had a very strong effect. We are currently attempting to dermine whether this is true for the other diseases as well. 4) An individual-level network model for a hypothetical outbreak of Japanese Encephalitis (JE) was developed for a selected scenario in the US. Simulation analysis suggested two important mitigation strategies: for low mosquito vectorial capacity, insecticidal spraying of infected areas reduces transmission and limits the outbreak to a single geographic area. Alternatively, in high mosquito vectorial capacity areas, birds rather than mosquitoes need to be removed/controlled. 5) Stochastic models were developed to study the spread of foot-and-mouth disease using actual farm number and farm sizes for the states of Oregon, Pennsylvania, and Texas. Results showed important effects, and that relationships differ substantially among the states. Preliminary results indicate the following: a. Varying the level of initial prevalence had a larger effect on epidemic expansion in Texas than in Pennsylania or Oregon. b. It appears that the farm size of the initial infections does nothave a big impact in OR or PA, but does increase epidemic size and duration somewhatin TX. c. Increased homogeneity of farm sizeincreased epidemic severity in Texas and Pennsylvania, but not Oregon. 6) A new, tangible geosptial modeling framework was developed for studying the spread of sudden oak death in heterogeneous landscapesand was applied to collaboratively managing disease spread. The following conclusions were drawn for the state of California: ". . . despite extensive cryptic (i.e., presymptomatic) infection and frequent long-range transmission, effective exclusion of the pathogen from large parts of the state could, in principle, have been possible were it to have been started by 2002. This is the approximate date by which sufficient knowledge ofP. ramorumepidemiology had accumulated for large-scale management to be realistic. The necessary expenditure would have been very large, but could have been greatly reduced by optimizing the radius within which infected sites are treated and careful selection of sites to treat. In particular, we find that a dynamic strategy treating sites on the epidemic wave front leads to optimal performance. We also find that "front loading" the budget, that is, treating very heavily at the start of the management program, would greatly improve control". 7) A modeling platform developed by the Kansas group was presented at the first investigator meeting, and picked up by the Oregon group to improve the computational efficiency of their model; this is an example of synergies expected from the collaborative approach to this project. 8) Three biological data sets collected at very different spatial scales werecomputationally combined for prediction of the dispersal of cereal rusts. This work now allows us to model the spread of infection from a single lesion to a distance of 10.5 kilometers, and is currently being used to evaluate the ability to model epidemics using data from different spatial scales. 9) Studies of dispersal kernels have shown kernels for wheat stripe ruststripe rust andfoot-and-mouth disease to be remarkably similar, despite large differences in the biology of these diseases. Studies are underway to make comparisons to kerenels of the other diseases.

Publications

• Type: Journal Articles Status: Published Year Published: 2016 Citation: Farber, D. H., Medlock, J. and Mundt, C. C. 2016. Local dispersal of Puccinia striiformis f. sp. tritici from isolated source lesions. Plant Pathology doi:10.1111/ppa.12554
• Type: Journal Articles Status: Published Year Published: 2016 Citation: McDonald, B. A., and Mundt, C. C. 2016. How knowledge of pathogen population biology informs management of Septoria tritici blotch. Phytopathology 106:948-955.
• Type: Journal Articles Status: Under Review Year Published: 2016 Citation: Farber, D. H., and Mundt, C. C. 2016. Effect of plant age and leaf position on susceptibility to wheat stripe rust. 2016. Phytopathology:in review.
• Type: Journal Articles Status: Under Review Year Published: 2016 Citation: M. Riad, C. Scoglio, L.W. Cohnstaedt. 2016. An individual-level network model for a hypothetical outbreak of Japanese Encephalitis in USA. Stochastic Environmental Research and Risk Assessment (SERRA):in review.
• Type: Journal Articles Status: Under Review Year Published: 2016 Citation: Tonini, F., Shoemaker, D., Petrasova, A., Harmon, B., Petras, V., Cobb, R.C., Mitasova, H., and Meentemeyer, R.K. 2016. Using tangible geospatial modeling for collaborative problem-solving: a pilot exercise with an invasive plant pathogen. Environmental Modeling:in review.
• Type: Journal Articles Status: Published Year Published: 2016 Citation: Cunniffe, N.J., Cobb, R.C., Meentemeyer, R.K., Rizzo, D., and Gilligan, C.A. 2016. Modeling when, where and how to manage a forest epidemic: sudden oak death in California. Proceedings of the National Academy of Sciences USA 113:56405645.