Epidemiology of Citrus Diseases · Epidemiology of Citrus Diseases Megan Dewdney PLP 5115c. What is Epidemiology? oThe study of epidemics Change in disease intensity in a host population

Epidemiology of Citrus Diseases

Megan DewdneyPLP 5115c

What is Epidemiology?oThe study of epidemicsChange in disease intensity in a host population over time and

spaceoChange: often an increaseDynamic process

oDisease: dealing with the ‘disease’, not just pathogen or crop (plant) Citrus canker rather than Xanthomonas citri subsp. citriHuanglongbing rather than Ca. Liberibacter asiaticus

What is Epidemiology? cont.

oHost: Organism (potentially) infected by another organismAlternaria Brown spot: Tangerines and tangerine hybrids

oPopulation: a population phenomena of both host and pathogenDynamic processes often described with statistics or

mathematical modelsoTime and Space: Two dimensions of interestChange over time or over a grove and sometimes both

Many Levels to Study Organisms

MolecularCellularTissueOrgan

IndividualPopulation

CommunitySystem

Epidemiology« Science of disease in

populations »(Vanderplank, 1963)

Broad DefinitionoEpidemic does NOT mean widespread or high levels of

diseasePandemic is the correct term for widespread or high levels of

diseaseoExample: Phytophthora infestans (Potato Late Blight)Field with 4 million plants (4 X 106)1 lesion/plant = 0.1% severity ~ 1/1000 leaf surface covered by

lesionsLimit of detection

LV Madden

Late Blight Example cont.

0

20

40

60

80

100

120

0 15 30 45 60 75 90 105 120

Dis

ease

seve

rity

(%)

Time (days)

% Severity

o time (t)=0 days(d) disease severity (y)=0.1% → t=90d y=100% 1000 fold change

o t=30d y=1% → t=90d y=100% 100 fold change

o t=0d y=1 lesion/field (0.1/4X106) → t=90d y=100% y=1 lesion/plant (0.1% severity or 1/4X106 lesions/field) – 4X106 fold change

Late Blight Example cont.oHow to determine when the epidemic started?oDoes scale change the biological processes that occur?oChange in population disease intensity is an epidemic

0

20

40

60

80

100

0 15 30 45 60 75 90 105 120

Dis

ease

seve

rity

(%)

Time (days)

% Severity

Disease Triangleo Ecology of diseaseo Principle of disease triangle still relevant but on

population levelEmphasis on interactions

o Time or space or humans or vectors?Awkward since limited to 3

dimensions

Francl, L.J. 2001. The Disease Triangle: A plant pathological paradigm revisited.The Plant Health Instructor. DOI: 10.1094/PHI-T-2001-0517-01; http://apsnet.org/education/InstructorCommunication/TeachingArticles/Francl/Top.html

Pathogen

EnvironmentHost

Epidemiology can be either…oDescriptiveWhere; when; whatHas been used to fill in disease cycles

ORoQuantitativeHow many ‘propagules’ are neededHow much disease is presentHow fast does disease developHow far can propagules travel

Tool Box

oClassical plant pathologyCulturing, microscopy, Koch’s postulates…

oTechniques from complimentary fieldsAgronomy, botany, ecology, entomology, genetics, statistics,

mathematics, meterology etc.

Host Growth and Susceptibilityo Melanose control requires good coverage with fungicide

on the fruit surface for nearly 3 monthso Copper is most common fungicideDoes not redistribute well on plant surfaceHas good residual activityCan build up in soilPhytotoxicity

o Foreseen problems?

Host Growth and Susceptibilityo Field study conducted to compare number of applications

with same amount of coppero More sprays reduced diseaseCovered up areas

on fruit exposed bygrowthLess wash off

Timmer et al, 1998

Host Growth and SusceptibilityoCopper residue can vary by year depending rainoModel developed to account for growth and rain

Timmer et al, 1998

Host Growth and SusceptibilityoWith no rain, copper residues will decline quickly

with rapid growth in early seasonoRain accelerates the processoMelanose cannot infect fruit > 8 cm dia.

No Rain Rain

oSize classes for fruit diameter1 = 20-25 mm, 2 = 26-35 mm, 3= 36-40 mm, 4= 41-60 mm, 5 > 60 mm

oLesion ratings0 = no lesions, 1 = discrete lesions within water-soaked (WS) area,

2 = coalesced lesion within WS area, 3 = coalesced lesion within and without WS area, 4 = expansion of lesions beyond 3 rating

Host Growth and Susceptibility

Graham et al, 1992

oCultivar susceptibility and age related or ontogenicresistance affects epidemic

oWhich fruit is most susceptible?oAs fruit become larger less susceptibleoTime is also a factor

Host Growth and Susceptibility

Graham et al, 1992

Any idea what disease this might be?

Host Growth and Susceptibility

o Why do fruit become more then less susceptible? Similar phenomenon in leaves

Stomates opening as fruit become larger?Xanthomonas citri subsp. citri may need expanding tissue to be

able to infectGrapefruit expands for longer during the season?

Surface waxes may not allow for as much wetting

Fruit Growth oDoes Grapefruit

expand for longer?

Stomates and Cankero It was thought that stomate size and density

would affect canker severityBut no relationship

oHost susceptibility on leavesOther factorsNot yet understood

Grapefruit

Cleopatra

Host Growth and Susceptibilityo Citrus leaves grow too fast to be effectively protected by

available fungicidesPyraclostrobin, copper hydroxide, ferbamExample is the case of Alternaria brown spotSimilar for Melanose and Citrus Scab Mondal et al., 2007

% D

iseas

eco

ntro

l

S = sprayedI = inoculated% = increase of leaf area between 2 dates

% D

iseas

eco

ntro

l

Environment

EnvironmentoCan affect whether a pathogen will infectAlternaria alternata and Xanthomonas citri subsp. citri

cannot infect if it is dryoPathogen dispersal is affected by environmentDiaporthe citri conidia are distributed by rain

oEnvironment influences inoculum productionMycosphaerella citri pseudothecia require wetting and

drying cycle to be initiated and matureoOther examples?

WindoTricky to work with in lab!

Gottwald and Graham, 1992

Regulated air supplyInoculum

Water for ‘Rain’

Effect of Wind on Canker

oThis is how it was determined that 8 m/s (18 mph) of wind driven rain was needed to force X. citri subsp. citri cells into a leaf

oLeaf expansion was also importantWhy?

Effect of Wind on Canker

oPressure also affected number of bacteria in leaves

oWhat is the difference in the two leaf surfaces?

What Environmental Stimulus is Needed?

oMany environmental stimuli were tested to see when A. alternata spores were releasedInside artificial chamber

Timmer et al., 1988

Environmental Stimuli cont.oRain and drops in relative humidity are not clearly

distinguishable but both contribute to spore releaseoIn field conidia production and infection weakly associated

with leaf wetness duration

Timmer et al., 1988

When are Conidia Produced?oField spore trapping of Pseudocercospora angolensisRelationship with temperature and rainfall more evidentSimilar pattern with relative humidityInteractions between variables not tested

Pretorius, 2005

Infection Conditions Alternaria Brown Spot

oOptimum temperatures 23-27°CCan get infection between 17-32°C

oInfection can occur with as little as 4-6 hours of leaf wetness but disease severity increases with leaf wetness

oAre there other factors that could affect this relationship?

Canihos et al., 1999

Infection Conditions Complicated by Hosto Not all cultivars react to the same infection conditions

identicallyAll susceptible hostsNova needs > 30 hours of leaf wetness to have same level of infection

as Minneola at 15 hours

0.00.10.20.30.40.50.60.70.80.91.0

2 4 6 8101214161820222426283032

01

23

45

Pro

port

ion

Leaf wetness (hours)

Rating

scale

20 C

0 lesions/leaf1-2 lesions/leaf3-5 lesions/leaf6-10 lesions/leaf11-15 lesions/leaf>15 lesions/leaf

0.00.2

0.4

0.6

0.8

1.0

2 4 6 8101214161820222426283032

0

1

2

3

4

5

Pro

port

ion

Leaf wetness (hours)Ra

ting

scale

20 C

0 lesions/leaf1-2 lesion/leaf3-5 lesion/leaf6-10 lesion/leaf11-15 lesion/leaf>15 lesion/leaf

0.00.10.20.30.40.50.6

0.70.8

0.91.0

2 4 6 81012141618202224262830323436384042

0

1

2

3

4

5

Pro

port

ion

Leaf wetness (hours)

Ratin

g sc

ale

20 C

0 lesions/leaf1-2 lesion/leaf3-5 lesion/leaf6-10 lesion/leaf11-15 lesion/leaf>15 lesion/leaf

Minneola NovaMurcott

Mondal et al. 2008

Probability of Disease

o Model developed from growth chamber data

o Prob.’s calculated for each lesion rating at the leaf wetness and temperature combination

P0 - No lesions

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30P1 - 1-3 lesions/leaf

20 22 24 26 28 30 32

5

10

15

20

25

30

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30

20 22 24 26 28 30 32

5

10

15

20

25

30

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95

Temperature (C)

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30

Temperature (C)20 22 24 26 28 30 32

5

10

15

20

25

30

P3 - 7-10 lesions/leafP2 - 4-6 lesions/leaf

P4 - 11-15 lesions/leaf P5 - greater than 15 lesions/leaf

Dancy - reduced model

Disease Probabilities cont.

oProbabilities change with cultivar

oSunburst is much less susceptible than Dancy

oReflected in graphs

P0 - No lesions

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30P1 - 1-3 lesions/leaf

20 22 24 26 28 30 32

5

10

15

20

25

30

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30

20 22 24 26 28 30 32

5

10

15

20

25

30

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95

Temperature (C)

20 22 24 26 28 30 32

Leaf

wet

ness

(hou

rs)

5

10

15

20

25

30

Temperature (C)20 22 24 26 28 30 32

5

10

15

20

25

30

P3 - 7-10 lesions/leafP2 - 4-6 lesions/leaf

P4 - 11-15 lesions/leaf P5 - greater than 15 lesions/leaf

Sunburst - reduced model

Lots of Interest in Leaf Wetness and Temperature

oConidia germinate6 hrs at 16 °C4 hrs 20 to 28 °C

oLiterature has varying times and temperatures needed for infection

oOptimum temp determined to be 24-28 °C

Agostini et al., 2003

Infection Conditions for Scab

oContradictory information in the literature about leaf wetness and temperature

oOptimal temperature range 23.5 to 27 °C

oOptimal leaf wetnessBetween 12 and 24 hrs

Agostini et al., 2003

Temperature Effect can Change with Disease Evaluationo Phytophthora palmivora - which disease?o What is the difference between incidence and severity? Incidence – disease status of plant units as individual or pieces such as

number of proportion of leaves with diseaseSeverity - area of disease

o How could this be important in an epidemic?

Timmer et al., 2000

Leaf Wetness and Temperature also Important for Inoculum Production

o Sporangia production highly dependant on both factorso Interaction also importantWhat is the significance

of an interaction?

Timmer et al., 2000

Pathogen EffectsoQuestions of interest about the pathogen:oWhat is required to produce inoculum?Are there environmental or other factors that contribute to

inoculum productionoHow much inoculum is present?Can affect how quickly an epidemic can become established and

move into exponential phasesoWhen is the inoculum present?No inoculum; no disease

Spore Traps

Burkard Spore TrapAllows for sampling spore patterns over time

Impact Traps/VolumetricAllows for sampling spores in a volume of air but not over time

• Spores are counted under the micro-scope

• Can be tedious and requires training

• Some new versions allow for PCR identification

Ascospore Ejection Patterno Phyllosticta spp. ascospore ejection is

reported to be triggered by raino In Brazil wetness duration was more

importanto Very frequent rain events; ascospores

cannot mature fast enough to eject with each rain event

o Cannot forecast infection event based on rainfall

Reis et al., 2006

Phyllosticta spp. Ascospore Release in FloridaWeek of May 13-20, 2010

Date

Thu 13 Sat 15 Mon 17 Wed 19 Fri 21

Num

ber o

f Gui

gnar

dia

asco

spor

es0

255075

100125150175200225250275300

Rain

(inc

hes)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Tem

pera

ture

(F)

65

70

75

80

85

90

95

Week of May 21-28, 2010

Date

Fri 21 Sun 23 Tue 25 Thu 27

Num

ber o

f Gui

gnar

dia

asco

spor

es

0255075

100125150175200225250275300

Rain

(inc

hes)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Tem

pera

ture

(F)

65

70

75

80

85

90

95

Pathogen PopulationsoHow many nurseries have metalaxyl resistant isolates of

Phytophthora nicotianae?oWhat proportion of the population?oIf nurseries have resistant isolates can spread around state

Timmer et al. 1998

Are Metalaxyl Resistant Isolates as Fit as Sensitive Ones?o Roots: similar proportion found as

added Resistant slightly more

o Soil: more resistant propagulesthan sensitiveMore propagules recovered than

applied (RYT)o Resistant strain more aggressive –

more likely to spreadTimmer et al., 1998

Bacterial DynamicsoVery few bacteria need to

penetrate leaves to initiate an infection

oIn 1 week have 107 cells in a lesion Many propagules formed!This is relatively slow for bacteria

Graham et al., 1992

Greasy Spot Inoculum Production

oWetting is critical for pseudotheciaproduction

oMost ascospores produced with the 3-day per week wetting scheme

oWetting scheme also changes peak ascospore ejection

Optimal Temperatures for Ascospore Productiono Spores trapped with a Burkhard trapSpores are produced within tight

temperature rangeSomewhat unusual but in Florida

conditions are within the optimal range often

Mondal and Timmer, 2002

Statistics and Mathematics

oMuch of epidemiology uses statistics especially the quantitative work

oMuch of the theoretical modeling that is undertaken uses a combination of mathematics and statistics

oA good working knowledge of statistics is needed to be a good epidemiologist and/or ecologistAt least know when to collaborate!

Disease Progress over Time

o Time is a fundamental factor in an epidemic since we are usually measuring change in disease status over timeNot a static processWhy some people include time in the disease triangle

o Often disease progress curves used to compare epidemics

Disease Progress of Canker Epidemic

oDisease progress curves at 5 urban sitesA is cumulative dataB is the rate of change between each

time pointoCan see this is a very dynamic

process as the rate of disease is not continuous

Gottwald et al, 2002Days

Epiphytic Growth and SeverityoGreasy spot severity is influenced by

when the epiphytic growth of Mycosphaerella citri occurs

oThe severity that occurs with levels of epiphytic growth changes over timeDisease severity does not track epiphytic

growth especially in the winter

Mondal and Timmer, 2003

Disease Progress in Space

oThere are two aspects of general interestDispersal gradientsSpatial patterns

oDispersal gradients tell how far an organism can spreadoSpatial patterns can give a sense of how the organism

spreadsSplash, wind, vector etc.Can indicate unforeseen dynamics in diseases

How Far Can A Sporangia Splash?

o Depends on speciesP. palmivora splashes further than P.

nicotianaeo Some strains travelled further than

otherso Means that P. palmivora is more likely

to move by splash and spread further

Timmer et al., 2000

Horizontal and Vertical Movement

oPhytophthora palmivora travels in 2 dimensions with water droplets

oAppears that majority of sporangia travel downGreater number of

colonies/sporangia below inoculum source

Timmer et al., 2000

Canker Frequency and DistanceoTried to find a distance where it was unlikely an infected

tree escaped579 m = 1900 ft

Gottwald et al, 2002

Common Spatial Patterns

UniformEvenly spaced patternUnusual in biological systemsSometimes from some sort of application mistake

AggregatedOccurs when the disease process depends on distance among individuals

RandomOccurs if disease process is independent of neighbors

How Many Samples Do I Need?o Want an accurate estimate of

pathogen populationo Need to know whether the pathogen

is commono From the patterns (with several

equations) arrived determined that: 1, 2, 3, 4, 5 or ten samples/tree were taken

then needed to sample 22, 13, 10, 8, 7 or 5 trees respectively

Timmer et al., 1998

More aggregated

Less aggregated

Urban Citrus CankeroWhat sort of pattern is this?oNote how few trees were affected initially

Gottwald et al, 2002

Citrus Scab Spread from a Foci

Gottwald, 1995

What Type of Spread Occurs with HLBo Wanted to know if spread

was from tree to tree in grove or from outside grove

o Used stocastic modeling to develop plots

MCMC posterior densitieso These plots show most

spread was mainly background (outside)

Gottwald et al, 2008

Spread Within Groveo Spread was mid-range distance so not spreading to

nearest neighbors but to nearby trees

Gottwald et al, 2008

What Kind of Spread is Occurring Here?

Gottwald et al, 2008

Spatial Patterns

oCould see with both Canker and Scab that the most likely trees to be infected were near by

oScab is splash distributedoCanker moves with wind-driven rainoAlso useful for understanding vectored diseasesThere is both external and medium range movement of infectious

Asian citrus psyllids

Disease Forecasting

o Two disease forecasting models used in citrusAlter-RaterPost-bloom Fruit Drop

o Designed so that the most effective timing of spray applications can be used

o Also predict decay of copper residue on fruit surfaces

ALTER-RATER: A Forecasting System

oWeather-based point system to better time fungicide applications

oPoints assigned based on:Rain fall and leaf wetnessAverage daily temperature

oThresholds vary by cultivar susceptibilityoHas been integrated into FAWN weather system

The ALTER-RATERSuggested Threshold Scores

Score Situation

50Heavily infested Minneola, Dancy, Orlando, Sunburst; Many flatwood groves, east coast, and SW Florida.

100 Moderately infested Minneola or Dancy, many Murcotts; Ridge and north Florida groves.

150 Light infestations, any variety, mostly Ridge and north Florida groves.

ALTER- RATER Daily PointsRain > 0.1 inch LW > 10 hr Avg daily Temp Assigned score

+ + 68-83 11+ + > 83 8+ + < 68 6+ _ 68-83 6+ _ > 83 4+ _ < 68 3_ + 68-83 6_ + > 83 6_ + < 68 4_ _ 68-83 3_ _ > 83 0_ _ < 68 0

Original PFD Model

577.1250048.016.163.13 ×+×++−= LWRTDy

y = Percentage of flowers infected 4 days in the future

TD = total number of infected flowers on 20 trees; however if TD < 75 then TD =0

R = rainfall total for the last 5 days in inchesLW = Average number of hours of leave wetness

daily for the last 5 days - 10 hours

When to Follow the Model

oA fungicide application is indicated if these three criteria are met:

1) the model predicts a disease incidence of greater than 20%2) sufficient bloom is present or developing to represent a significant portion of the total crop

3) no fungicide application has been made in the last 10-14 days.

How the Citrus Copper Application Scheduler Operateso Incorporates rainfall data from FAWN (Florida Automated

Weather Network-www.fawn.ifas.ufl.edu) or own weather data

o Incorporates data on copper residue degradation

o Incorporates fruit growth size

Steps to Achieve Daily Prediction

Zortea et al. (2012)

Series of EquationsoModel is built on series of equationsCopper application residue𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 = 0.6399 + 0.005539 V A ( C

4)

Fruit growth

𝐴𝐴𝐴𝐴𝐷𝐷𝐴𝐴 = 𝑀𝑀𝐴𝐴𝑀𝑀 × 𝑒𝑒ln(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀)𝑒𝑒−𝐵𝐵𝐵𝐵

Residue for each day𝐴𝐴𝐷𝐷𝑅𝑅𝑅𝑅𝐷𝐷𝑅𝑅𝐷𝐷 = 𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷

𝐴𝐴𝐴𝐴𝐷𝐷𝐴𝐴

Residue loss 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = 𝐴𝐴𝐷𝐷𝑅𝑅𝑅𝑅𝐷𝐷𝑅𝑅𝐷𝐷 ( 0.016535𝐴𝐴 )

Zortea et al. (2012)

To Use

oSelect weather option and scion first

oCan use metric units

Enter Bloom Date

Every 21- day Schedule

oHave insufficient coverage for 6 days

oAbout perfect timing for third spray

Coverage Optimized with ModeloMoved first spray

up 8 days

oDid not move third spray

Improvements in Progress

oSome operations cannot easily take advantage of modelEquipment movementNeed to schedule in advance

oDeveloped optimized schedule for such operationsHistorical weather per regionBloom date

Traditional Versus Optimizedo21-day schedule (top) had

2 major gaps in coverage

oOptimized schedule reduced gaps in coverage but did not eliminate

Further Improvements

o Original model not designed to predict past mid JulyWhy?

o Need residue data for summer

o What diseases?o Fruit growth too

Grapefruit

Date

Jun Jul Aug Sep Oct Nov

Tota

l dai

ly ra

infa

ll (m

m)

0

20

40

60

80

100

120

Ave

rage

cop

per/s

urfa

ce a

rea

( g/

cm2 )

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0