Karen Garrett - Managing Pests and Diseases Under Climate Change

1,035 views
794 views

Published on

Presented at a video science seminar on 18 April 2012. Karen Garrett is a Professor in plant pathology at Kansas State University. Research conducted under the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). Watch at http://ccafs.cgiar.org/events/18/apr/2012/live-video-science-seminar-managing-pests-and-disease-under-climate-change-what

Published in: Education, Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
1,035
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
31
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Karen Garrett - Managing Pests and Diseases Under Climate Change

  1. 1. Managing pests and diseases under climate change: What do we need to learn? Karen Garrett Kansas State University kgarrett@ksu.edu
  2. 2. Outline• Inputs needed for a system of adaptation• ‘Simple’ estimates of climate change effects on disease/pest risk• Identifying system components for inclusion in climate change scenario analyses• Impacts of system variance and the color of weather time series
  3. 3. Yield loss (Oerke 2006)
  4. 4. Scoping study for pests/diseases• Priorities to address pests and diseases in agricultural adaptation to climate variability and change: East Africa, West Africa, and the Indo-Gangetic Plains• K. Garrett, Andy Dobson (Oxford), Juergen Kroschel (CIP), Simone Orlandini (U Florence), M. Sporleder (CIP), Henri Tonnang (CIP), Corinne Valdivia (U Missouri)• We are selecting crop and livestock systems to emphasize based on their importance and likely response to climate change, and welcome your input
  5. 5. 1. Weatherand climatedata as inputinformation Hourly/daily/weekly weather Seasonal/annual weather Decadal climate
  6. 6. 2. Input intodisease/pest modelwith appropriatespecies and locationspecificity 2d. Specific geographic 2a. Generic 2b. Specific location: Base disease/pest disease/pest: Base information/models platform for information/models interface with 2e. Specific 2c. Specific crop models and geographic disease/pest: weather data location: Iteratively Iteratively improved improved information/models information/models
  7. 7. 3. Disease/pestanalysis toinform decision-makers 3b. Farmers’ species and variety/breed 3d. 3a. Farmers’ decisions Donors’/scientists’ / chemical, cultural, and policy makers’ and biological prioritization decisions 3c. Insurance rate decisions and payout decisions
  8. 8. 4. Bettermanagement toimprove systemoutcomes Improved food security, livelihood, and environment
  9. 9. Analysis of ‘simple’ temperature and relativehumidity effects may provide a ‘first orderapproximation’ to the future level of riskWe used models of potato late blight riskdeveloped by Fry et al. and Grunwald et al.,rescaling them for use with coarser temporalresolution data at larger spatial extent
  10. 10. Estimated Potato Late Blight Risk For current potato production regions Adam Sparks, R. Hijmans, G. Forbes, K. Garrett1961-1990
  11. 11. Estimated Potato Late Blight Risk For current potato production regions Adam Sparks, R. Hijmans, G. Forbes, K. Garrett2040-2060
  12. 12. Complexity in terms of the amount ofinformation needed to adequatelypredict outcomes
  13. 13. Lower complexity Higher complexity1. Are multiple A single pathogen Microbialbiological species ‘acting alone’ communities, vectorinteractions causes disease in a communities, and/orimportant? single plant species complex landscapes influence disease outcomes Fusarium head blight Potato late blight ++ + Multiple host species Multiple host species Multiple pathogen species
  14. 14. Lower complexity Higher complexity2. Are there Pathogen population Pathogen populationenvironmental responses to climate responses changethresholds for variables are constant suddenly atpopulation throughout the particular thresholdsresponses? relevant range Fusarium head blight Potato late blight 0 0 Little evidence for Little evidence for this this
  15. 15. Allee effects• Disproportionate reduction in reproductive success at low population densities• Mechanisms – Difficulty finding a mate -> Karnal bunt of wheat – Lack of predator saturation – Impaired group dynamics – Environmental conditioning
  16. 16. Lower complexity Higher complexity3. Are there indirect The relationship Global change factorseffects of global between climate (land use, water,change factors on variables and disease transportation,disease risk is unrelated to markets) influencedevelopment? other factors the relationship Fusarium head blight Potato late blight + + Land use change: Enhanced transport maize and wheat co- networks: greater occurrence, and exchange of seed and reduced tillage pathogen systems populations
  17. 17. Lower complexity Higher complexity4. Are spatial Disease risk at a given Climate maycomponents of location is not influence theepidemic processes influenced by disease likelihood of diseaseaffected by climate? risk at other locations spreading among locations Fusarium head blight Potato late blight + + Temporal/ Aerobiology/ phenological dispersal may be requirements for modifed infection
  18. 18. MaizeRed = connected areas forpathogens that require atleast low maize density tospread Red = connected areas for pathogens that require high maize density to spread
  19. 19. Lower complexity Higher complexity5. Are there feedback Management tools Management efficacyloops for have the same level changes greatly withmanagement? of efficacy despite changes in the changes in other system system components Fusarium head blight Potato late blight ++ + Worldwide, as Buildup of inoculum sanitation and other from multiple host techniques may species makes become less useful management more ++ Highland tropics, difficult as colder highlands disappear
  20. 20. IRRI
  21. 21. The benefits of potato mixtures for late blightmanagement were lower in areas with longergrowing seasons (Cajamarca-Peru, Quito-Ecuador) compared to areas with shortergrowing seasons (Huancayo-Peru, Corvallis-USA)Where season length increases under climatechange, methods such as resistancedeployment in mixtures (and sanitation,certain resistance genes) may become lessuseful
  22. 22. Lower complexity Higher complexity6. Are networks for Disease is already Disease moves tointervention present in relevant new areas wheretechnologies slower areas, or is well farmers do not havethan epidemic understood and tools and knowledge;networks? easily managed no knowledge networks Fusarium head blight Potato late blight + ++ Online risk Reliance on evaluations available pesticides, with high for some regions knowledge requirements; Slower vegetative propagation of resistant varieties
  23. 23. Geo node 2 Geographic network Geo node 1 Microbial communities Plant communities & plant disease status Human information Environment Geo node 3
  24. 24. Lower complexity Higher complexity7. Are there effects Disease has impacts Disease may impactof plant disease on only on yield of a other plant speciesmultiple ecosystem single host plant and/or health ofservices? species humans, other animals, and soils Fusarium head blight Potato late blight +++ + Management may Increased LB results increase erosion and in increased pesticide reduce support for exposure for humans wildlife; mycotoxins and environment create health risks
  25. 25. Biological impactsPolicy from beyond target region Ecosystem services Resistance genes Provisioning services Invasive species e.g., Food Supporting services e.g., Nutrient cycling Regulating services e.g., Climate regulationSystem diversity Pests/disease Cultural services e.g., Recreation & their management 2009 Phytopathology 99:1228-1236.
  26. 26. Lower complexity Higher complexity8. Are there feedback Disease is affected by Epidemics affectloops from plant climate change, but climate changedisease to climate has no impact on factors such as soilchange? climate change erosion and/or photosynthetic capacity Fusarium head blight Potato late blight + + Increased ‘carbon Increased ‘carbon cost’ of wheat cost’ of potato production; Tillage production may reduce carbon sequestration
  27. 27. The effects of climate variability and the color ofweather time series on agricultural diseases andpests, and on decisions for their management• K. Garrett, Andy Dobson (Oxford), Juergen Kroschel (CIP), Bala Natarajan (KSU), Simone Orlandini (U Florence), Henri Tonnang (CIP), Corinne Valdivia (U Missouri)• Agricultural and Forest Meteorology, in revision
  28. 28. In this theoretical model, yield loss grows following a logistic curve, with the rate of yield loss determined by the time series of weather conduciveness to yield lossThe first results shown are in the absence of management
  29. 29. Color of disease/pest Disease/pest loss Yield loss rate Cumulativeloss conduciveness conduciveness time series yield losstime series time series time series White: no autocorrelation Light pink: Moderate v autocorrelation Darker pink: High v autocorrelation
  30. 30. Color of disease/pest Disease/pest loss Yield loss rate Cumulativeloss conduciveness conduciveness time series yield losstime series time series time series White: no autocorrelation Light pink: Moderate autocorrelation Darker pink: High autocorrelation
  31. 31. Color of disease/pest Low mean Intermediate mean High meanloss conduciveness conduciveness conduciveness conducivenesstime series to loss to loss to loss Yield Loss Yield Loss Yield Loss White: no autocorrelation Variance Variance Variance Yield Loss Yield Loss Yield Loss Light pink: Moderate autocorrelation Variance Variance Variance Yield Loss Yield Loss Yield Loss Darker pink: High autocorrelation Variance Variance Variance
  32. 32. Color of disease/pest Low mean Intermediate mean High meanloss conduciveness conduciveness conduciveness conducivenesstime series to loss to loss to loss Yield Loss Yield Loss Yield Loss White: no autocorrelation Variance Variance Variance Yield Loss Yield Loss Yield Loss Light pink: Moderate autocorrelation Variance Variance Variance Yield Loss Yield Loss Yield Loss Darker pink: High autocorrelation Variance Variance Variance
  33. 33. Resulting hypothesesIn the absence of management:Under less conducive mean conditions, increasing variance increases yield lossUnder more conducive mean conditions, increasing variance decreases yield loss
  34. 34. What information will farmers use for decision-making?• Past years – In our simulation: If yield loss exceeds the cost of management in 2 out of 3 years, use management this year• Current year – In our simulation: If yield loss has begun by mid- season (>5%) and is not too high (< 100% minus cost of management), use management this year
  35. 35. Resulting hypotheses for success of decision-making• False positive: decision to manage when it reduces profit• False negative: decision not to manage when it would have increased profit• Low or intermediate mean conduciveness to pest/disease loss: – Increasing variance increases false positives and false negatives – Higher temporal autocorrelation decreases false positives and false negatives• High mean conduciveness to pest/disease loss: increasing variance decreases false positive rates
  36. 36. Outline• Inputs needed for a system of adaptation• ‘Simple’ estimates of climate change effects on disease/pest risk• Identifying system components for inclusion in climate change scenario analyses• Impacts of system variance and the color of weather time series
  37. 37. kgarrett@ksu.edu Thanks for your attention! Thanks to all the collabo rators on these projects Support provided by US NSF, US DOE, USDA, USAID, CGIAR Photo: Pete Garfinkel
  38. 38. END OF MAIN PRESENTATION• Additional slides are supplements that might be useful during the discussion
  39. 39. What do we need to understand to adapt to climate change?• For growers – How to adapt early warning systems for within-season tactical decision making – How to construct longer-term (season or longer) support for decision making• For plant breeders – What diseases to prioritize where• For policy makers / donors – What the important disease problems are for investment in the future – How financial tools can buffer farmers from increased variability• In natural systems – A lot… including the distribution of resistance genes
  40. 40. Crop arthropod pests• Potato tuber moths: Phthorimaea operculella, Tecia solanivora, Symmetrischema tangolias• Potato leafminers: Liriomyza huidobrensis, L. trifolii, L. sativae• Cruciferae: Plutella xylostella• Maize: Plutella xylostella• Sweetpotato weevils: Cylas puncticollis and C. brunneus• Sweetpotato butterfly: Acraea acerata• Potato whitefly: Bemisia tabaci Type A and B. afer, Trialeurodes vaporarium• Maize stem borers: Chilo partellus, C. orichalcociliellus and Busseola fusca; Sesamia calamistis• Cassava mealy bug: Phenacoccus manihoti• Cassava green mite: Mononychellus tanajoa• Mango, Sri Lanka fly: Bactrocera invadens
  41. 41. Crop diseases• Maize streak virus• Rice blast, brown spot, grain rot, and bacterial blight• Pigeonpea Phytophthora blight• Wheat rusts• Potato late blight• Virus complexes of cassava, sweetpotato, and potato• Chickpea dry rot (Rhizoctonia bataticola)• Banana bunchy top virus, Xanthomonas wilt, and Panama disease• Mycotoxin-producing pathogens
  42. 42. Livestock pests and diseases• Fascioliasis: wide host range, incl. humans• African Horse Sickness: Horses, mules, donkeys• Rift Valley Fever: Livestock and humans• Peste de Petites Ruminants (PPR): goats and sheep
  43. 43. Sparks, Forbes, Hijmans, Garrett, 2011
  44. 44. Sparks, Forbes, Hijmans, Garrett, 2011
  45. 45. Conducive year every 3 years 9 8 Log(Viable teliospores at beginning of year) 7 6 5 4 c 3 2 Allee ef fect 1 No Allee eff ect 0 1 3 5 7 9 11 13 17 19 15 YearGarrett and Bowden 2002 Phytopathology
  46. 46. Conducive year every 4 years 9 8 Log(Viable te liospores at beginning of year) 7 6 5 Allee effect 4 No Allee effect 3 2 1 0 1 3 5 7 9 11 13 15 17 19 YearGarrett and Bowden 2002 Phytopathology
  47. 47. Color of disease/pest Low mean Intermediate mean High meanloss conduciveness conduciveness conduciveness conducivenesstime series to loss to loss to loss White: no autocorrelation Proportion Incorrect Decisions Proportion Incorrect Decisions Proportion Incorrect Decisions Variance Variance Variance Light pink: Moderate autocorrelation Variance Variance Variance Dark pink: High autocorrelation Variance Variance Variance
  48. 48. Color of disease/pest Low mean Intermediate mean High meanloss conduciveness conduciveness conduciveness conducivenesstime series to loss to loss to loss White: no autocorrelation Proportion Incorrect Decisions Proportion Incorrect Decisions Proportion Incorrect Decisions Variance Variance Variance Light pink: Moderate autocorrelation Variance Variance Variance Dark pink: High autocorrelation Variance Variance Variance

×