next generation crop insect pest models for integrated ... · simon (2013): predicting climate...

The Insect Life Cycle Modeling (ILCYM) software and options for

integrated assessments

AG

MIP

, Feb

ruar

y, 2

015

J. KroschelM. Sporleder, P. Carhuapoma, H. Juarez, N. Mujica

DCE Crop Systems Intensification and ClimateChange (CSI-CC); Agroecology/IPM sub-program

Next generation crop insect pest models for integrated

assessments

Outline

� Introduction

• Development of ILCYM: For which purpose

• Main features and advantages

� ILCYM applications

• Pest Risk Analysis (PRA) on different scales

• Climate change and adaptation planning

• Regional assessments

• Decision support tool in pest management

�Outlook: considerations in integrated assessments

� Conclusions

1 Species distribution mapping (SDM) in support of pest risk

assessment (PRA, IPPC)

2 Climate change / adaptation planning

3 Integrated Pest Management (Examples)

- Simulation of pest life-table parameters under prevailing weather

conditions as decision support for pest management

- Identification of potential release sites for parasitoids (classical biocontrol)

- Simulation of field performance of biopesticides and application rates

Analytical tools for predicting, evaluating and understanding the dynamics of

insect populations in agroececosystems.

ILCYM: For which purpose

Potato Tuber Moth, Phthorimaea operculella

• Invasive species; reported from more than 90 countries

• Adapted to wide range of different climates and agroecologies

• Yield losses due to leaf and tuber infestation at harvest

• Storage pest in tropical and subtropical countries

Potato production areas and PTM distribution

Phthorimaea operculella

Symmetrischema tangolias

Tecia solanivora

Likelihood of entry and

establishment*Consequences of

establishment

Entry potential

Colonization potential

Spread potential

Economic impact potential

Environmental impact potential

Social impacts

1

2

3

A B

Establishment potential

Framework for Pest Risk Analysis

*under current and future climates

Insect Life Cycle Modeling (ILCYM)

• Required input: life-table data collected under

constant and fluctuating temperatures

• Model builder:

- Fitting functions for describing temperature-

driven development processes under constant

temperatures, i.e. development rate, mortality

and reproduction

• Validation and simulation module

- Uses life-table data established under

fluctuating temperature; can be used for

deterministic and stochastic simulations

• Potential population and risk mapping module

- Establishment index

- Generation index

- Activity index

Software package for developing temperature-based insect phenology models with

applications for regional and global pest risk assessments and mapping.

(www.cipotato.org/ilcym)

Pest risk mapping using ILCYM

Model validation

Temperature-based phenology

modelsMo

de

lb

uil

de

r

Development rate and time Survival rate, oviposition, mortality

Risk indices

ERI GI AI

Ris

km

ap

pin

g

ERI= ∑ �� /� ��

∑ 365� �� /�� 365GI=

log�� !"��

��AI=

Potential distribution at global, regional

and local level

Future climate: 2050With and without crop shapes

Current climate: 2000(1950-2000: www.worldclim.org/) Down-scaled data SRES-A1B, IPCC (2007):

http://gisweb.ciat.cgiar.org/GCMPage

(www.cipotato.org/ilcym)

Simulation of life-table

parameters

Net reproduction rate, generation time

0 10 20 30

50

100

50

100

150

0 10 403020 0 10 20 30

0.00

0.08

0.04

1.25

1.15

1.05

0.95

10 20 30 40

0

150

100

50

0 10 20 30 40

intrinsic and finite rate of increase, double time

Life-table studies

A constant and fluctuating

temperatures: oviposition,

survival time, development time,

mortality

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60

Indi

vidu

als

0

5

10

15

20

25

30

35

40

Tem

pera

ture

(°C

)

Arequipa 2do cycle, 13.01.99

Larva

Pupa

Eggs

Adult

Time (days)

(Keller 2003)

Model validation

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100 110

Indi

vidu

als

0

5

10

15

20

25

30

35

40

Tem

pera

ture

(°C

)

Huancayo 2do. cycle, 24.11.98

LarvaePupae Adults

Eggs

Time (days)

(Keller 2003)

Model validation

Establishment risk index (Survival risk index): Indicates the capacity of

invasive pest species to establish permanent populations

Index =

The index is defined as the number of time intervals with a net reproduction rate, R0, above 1 (Ii=1)

divided by the total number of time intervals within a year (II).

Generation Index: Indicates the potential infestation of a crop

Index =

The index is computed by averaging the sum of estimated generation lengths, Ti , calculated for

each Julian day, x, where 365 is the number of days per year and Ti is the predicted average

generation lengths in days during the month, i (i = 1, 2, 3, …, 12).

Activity Index: Indicates pest potential population growth, spread and severity

Index =

where λλλλ is the finite rate of increase at Julian day, x (x = 1, 2, 3, …365).

Three risk indices

∑ #$� �� /365%&'(#$� �)

∑ 365�(�� /�� 12

log�� !"��

��

Main advantages of ILCYM modeling (1)

• Complex simulation of an entire insect life system.

• It uses process-based functions that provide a mechanistically detailed

representation of the insects life history, and hence population growth.

• Model functions are build based on statistical analysis of experimental

data that allows determining errors in particular processes.

• Full live history (development, survival, reproduction) is modeled and

the software compiles the overall model automatically (reducing

programming skills and time of users)

• Models have parameters with straightforward interpretation of the

biological system.

• Individual functions of a model can be validated and improved

• The information can be used in different modeling types; i.e. spatially

(SDM through the use of specific risk indices), or in time (in a given

area) for different inquiry (from basic ecological understanding to

prediction, management and policy-making).

Main advantages of ILCYM modeling (2)

• Since the model works with probability distributions for specific

processes it is possible to conduct both deterministic and stochastic

simulations.

• The simulation uses an cohort-updating algorithm and quantifies the

population age-stage specific.

• Input data can be real or simulated temperature at variable intervals.

• The simulation is implemented in R-language, which is open source

software.

ILCYM: Reference publications

Sporleder, M., H.E.Z. Tonnang, P. Carhuapoma, J. C. Gonzales, H. Juarez & J. Kroschel

(2013): Insect Life Cycle Modeling (ILCYM) software – a new tool for Regional and

Global Insect Pest Risk Assessments under Current and Future Climate Change

Scenarios. CABI Book publication, 412-427.

Tonnang, E.Z.H., P. Carhuapoma, H. Juarez, J.C. Gonzales, D. Mendoza, M. Sporleder, R. Simon,

J. Kroschel (2013): ILCYM - Insect Life Cycle Modeling (version 3.). User Guide. A software

package for developing temperature-based insect phenology models with applications for

regional and global analysis of insect population and mapping. International Potato Center,

Lima, Peru. pp 193.

Sporleder, M., J. Kroschel, M. R. Gutierrez Quispe & A. Lagnaoui (2004): A temperature-

based simulation model for the potato tuber worm, Phthorimaea operculella Zeller

(Lepidoptera; Gelechiidae). Environmental Entomology 33 (3): 477-486.

Kroschel, J., M. Sporleder, H.E.Z. Tonnang, H. Juarez, P. Carhuapoma, J.C. Gonzales & R.

Simon (2013): Predicting climate change caused changes in global temperature on potato

tuber moth Phthorimaea operculella (Zeller) distribution and abundance using phenology

modeling and GIS mapping. Journal of Agricultural and Forest Meteorology 170: 228-241.

ILCYM users

Number of ILCYM potential users that have visited the software webpage from

March 1, 2010 until 20 March 2014

Pest Risk Mapping and

new framework to assess

climate change impacts!

Rational for

phytosanitary

requirements

(by NPPO)

Stage IV:

Documentation

Stage IV:

Documentation

Pest identification

Pathway

Policy review

PRA area

Pest taxonomy etc.

Stage I:

Initiation

Stage I:

Initiation

Pest categorization

Probability

X

Magnitude of

econ./environ.

Impact

Stage II:

PR-Assessment

Stage II:

PR-Assessment

*(adopted from FAO-IPPC, 2007)

React

Test & evaluate

efficacy of

measures

Surveillance

Diagnostics

Stage III:

PR-Management

Stage III:

PR-Management

Applications of ILCYM: Pest Risk Analysis (PRA)

“The process of evaluating biological or other scientific and economic

evidence to determine whether an organism is a pest, whether it

should be regulated, and the strength of any phytosanitary measures

to be taken against it”

National and regional

adaptation planning!

Pest Risk Mapping

Validation of risk maps by using geo-referenced distribution maps

Example: potato tuber moth

Green points: countries with reported pest establishment

Red points: geo-referenced distribution data

Yellow points: permanent establishment not confirmed

ILCYM: Risk mapping under current and future

climates Phthorimaea operculella: ERI 2000 and 2050

A range expansion into temperate regions of the northern hemisphere and in

tropical mountainous regions

(Kroschel et al., J. Agric. Forest Meterol., 2013)

Europe:

2.73% (234,404 ha) North America:

4.21% (32,873 ha)

Asia:

8.3% (699,680 ha)

Africa:

1.01% (11,605 ha)

Oceania:

13.67% (7,314 ha)

South America:

11.9% (109,666 ha)


climates


Phthorimaea operculella: GI 2000 and 2050Abundance and damage potential will progressively increase in all regions

where the pest already prevails today with an excessive increase in warmer

cropping regions of the tropics and subtropics


climates Phthorimaea operculella: GI Change

>4 generations are developed on 30.2% (5.918.557 ha) of the total potato

production area worldwide; this can pot. increase to 42% (8.328.531 ha)


Africa:

2.87% (1,113,966 ha)

Asia:

15.29% (5,246,319 ha)Europe:

10.14% (1,115,432 ha)

North America:

15.87% (280,180 ha)

Oceania:

26.05% (42,465 ha)

South America:

13.11% (530,116 ha)


climates P. operculella validation: observed and predicted numbers of

generations at different locations using different climate input data

Country/location

Life table

studies under

local natural

conditions

ILCYM modeling

results using local

weather station

data

ILCYM modeling

results using

WorldClim data

Monthly records

Peru: San Ramon,

800 m a.s.l.12 12 9.5

Peru: Arequipa,

1140 m a.s.l.71 6.7 7.2

Peru: Huancayo,

3250 m a.s.l.3-4 3.1 2.3

Yemen: Sana’a,

2150 m a.s.l. 82 7.1 5.4

Egypt: Giza,

10 m a.s.l.103 10 8.6



climates Effect of climate change on the number of generations at different locations

Nu

mb

er

of

ge

ne

rati

on

s p

er

ye

ar

0

2

4

6

8

10

12

14

16

18

Peru - San Ramon (800 m) Peru - Huancayo (3250 m) Peru - Arequipa (1440) Yemen - Sana’a (2250) Egypt - Giza (10 m)

12.5

12

+1.5

+2.5

+3.9+4.6

2.9+0.4

+0.7+1.0

+1.2

6.7+0.8

+1.4+2.4

+2.8

3.0

7.0

8.0

7.0

+0.9+1.3

+2.2+2.7 10 10

+1.2+2.0

+3.7+3.1

PRESENT PREDICTION PRESENT PREDICTION PRESENT PREDICTION PRESENT PREDICTIONPRESENT PREDICTION

Leafminer fly, Liriomyza huidobrensis

• Center of origin in the neotropics; reported from more than 66 countries

• Very polyphagous; in Peru pest in >27 horticultural crops

• Yield loss without management up to 70% in potato

Green points: countries with reported pest establishment

Yellow points: reported in protected crops

Red points: geo-referenced distribution data


climates Liriomyza huidobrensis: Regional maps

ERI

GI

2000 2050Index change:

2000-2050

2000 2050 Change

Act

ivit

y i

nd

ex

Ge

ne

rati

on

in

de

x


climates Liriomyza huidobrensis: country maps for Uganda

L. huidobrensis will potentially increase its risks in the northern, eastern and

southern provinces due the high values of GI>12 and AI>17

Publication Project – Pest Risk Atlas

Pest Distribution and Risk Atlas for AfricaPotential global and regional distribution and abundance of

agricultural and horticultural pests and associated biocontrol agents under current and future climates

Potato pests Potato tuber moth, Phthorimaea operculella (Kroschel et al.)Guatemalan potato tuber moth, Tecia solanivora (Schaub et al.)Andean potato tuber moth, Symmetrischema tangolias (Sporleder et al.)

Sweetpotato pestsSweetpotato weevil, Cylas puncticollis (Okonya et al.)Sweetpotato weevil, Cylas brunneus (Paul Musana et al.)Sweetpotato butterfly, Acraea acerata (Okonya et al.)Sweetpotato white fly, Bemisia tabaci Type B (Gamarra et al.)White fly, Bemisia afer (Gamarra et al.)

Vegetable pestsSerpentine leafminer fly, Liriomyza huidobrensis (Mujica et al.)Vegetable leafminer, Liriomyza sativae (Mujica et al.)American serpentine leafminer, Liriomyza trifolii (Le Ru et al.)Greenhouse whitefly, Trialeurodes vaporarium (Gamarra et al.)

3.4 Maize pestsSpotted stemborer, Chilo partellus (Khadioli et al.)Maize stalk borer, Busseola fusca (Khadioli et al.)African pink stem borer, Sesamia calamistis (Khadioli et al.)

3.5 Cassava pestsCassava mealybug, Phenacoccus manihoti (Hanna et al.)Cassava green mite, Mononychellus tanajoa (Hanna et al.)

3.6 Fruit pestsOriental fruit fly, Bactrocera invadens (Hanna et al.)Banana aphid, Pentalonia nigronevosa (Hanna et al.)

Publication Project – Pest Risk Atlas

Reducing crop vulnerability by adaptation to new and emerging pests

+ Adaptation (of IPM)

Resilient systems, new technologies,

tolerant and resistant varieties reduce pest

caused losses by 15%

Exposure

Temperature

increase: 2 °C

Vulnerability

Actual impact on crop yields by pests increase by 5%

Mitigation Reduction of CO2 emissions

(Adapted from IPCC, 2001, 2007)

Crop yields Crop pests

Sensitivity

Pot. impacts on crop yields based on pest

expansion/increased abundance: Higher

infestations reduce yields by 20%

Climate change / adaptation planning

ObjectiveTo inform stakeholders in plant health,plant quarantine and pest managementabout future pest risks on global,regional and local scales and to usethis information for crop-specificadaptation planning and to definepriorities in pest management.

Stakeholder workshop

Climate change / adaptation planning

Climate data and map resolution

• For estimating pest risk for specific locations, max. and min.

temperature data in a 1 hr time step length are used to

consider the within-day temperature variability that may

affect insects.

• For global and regional pest risk mapping, currently

available temperature surfaces use monthly average

aggregated climate data from 1960 to 2000

(www.worldclim.org).

• Aggregation and interpolation of temperature data has its

limitations especially in mountainous regions due to local

temperature variability (topographic effects) over short

distances.

• Higher accuracy of risk maps and predictions of potential

pest distribution could be achieved through locally collected

daily weather data processed and interpolated in ILCYM.

0

5

10

15

20

25

30

35

0 4 8 12 16 20 24

Time steps

Tem

pera

ture

Min

Min

Max

� Development of a tool which allows regional assessments of pest risks in

mountainous regions at high resolution.

ILCYM: regional assessments

18 x 18km (1x)

Map resolutions: from global to regional and local scales


18 x 18km (1x)9x9km (4x respect to 10’)



18 x 18km (1x)9x9km (4x respect to 10’)4.5x4.5km (16x respect to 10’)



18 x 18km (1x)9x9km (4x respect to 10’)4.5x4.5km (16x respect to 10’)0.9x0.9km (400x respect to 10’)



18 x 18km (1x)9x9km (4x respect to 10’)4.5x4.5km (16x respect to 10’)0.9x0.9km (400x respect to 10’)

90m x90m (40,000x respect to 10’)



21 weather stations

were installed in the

Mantaro Valley, Peru

Data collection


Climateinterpolator(thin plate smothing spline)

[Daily]Tmax n=365, Tmin n=365

[Mapping Index: GI, ERI, AI]

[Calculating theIndices (GI, ERI, AI) per weatherstation]

GI, ERI, AI

IndexInterpolatortool(thin platesmothing spline)

ANUSPLIN 4.1 (Australia)Data processing:

30 days versus 5 min.


Local risk maps for leaf miner fly, Mantaro Valley, Peru

ERI GI

Use of ILCYM’s Index Interpolator



ERI+1 °C GI+1 °C




ERI+2 °C GI+2 °C




ERI+3 °C GI+3 °C



RTB-Pest Risk Assessment project

Impact of climate change on the

livelihood of farmers

Household models

RTB yield

models

Pest and crop models

(yield gaps caused by insect pest)

Modeling climate change

New entry

pests (PRA)

Current

pests


(Framework of RTB Pest Risk Analysis (PRA) project)

Burundi Rwanda

0

500

1000

1500

2000

2500

3000A

ltitu

de (

m a

sl)

Weather stations

Action sites of RTB PRA project:

Burundi, Rusizi valley: 900 to 2600 m asl

Rwanda, Ruhengeri: 1400 to 2600 m asl

- All RTB crops

- Harmonization and standardization

of household and field survey

methodologies and collection of climatic

information along an altitudinal gradient


Pest management (1)

1. Use of calendar date: Pest status assessed at

a specific date or crop growth stage and

pesticides are applied if the pests economic

threshold is reached.

Example: potato tuber moth

Control threshold: 0.5 mines/plant

Critical infestation period: GS 60

2. Use of degree-days (DD): The temperature

above the pest’s specific temperature

minimum is summed up (degree-day

summation) and the pest status assessed

when the pest specific temperature sum is

reached (improvement to 1).

Example: Evolution in decision support tools

Determination of pesticide application date

Mines / plant

Example:

Daily pest degree

day calculator of

the Illinois State

Water Survey,

University of

Illinois

3. Use of non-linear models for insect development: (daily development rates are

summed up until reaching a value of 1 (further improvement to 2, specially

when temperature is in the non-linear range for development).



Pest management (2)

4. Process-based climatic response model (ILCYM phenology model): Temperature

variability in autumn influences emerging time of insects in spring because the

overwintering stage already past some development in autumn (Example: mountain pine

beetle, Dendroctonus ponderosae (Hicke et al. 2006).

Huancayo, Peru (3250 masl) Hermiston, Oregon, USA (144 masl)

Example using ILCYM: Potato tuber moth: Rate of population increase within one year



Pest management (3)

Highland agroecology (3250 masl)

Simulation of pest life-table parameters under prevailing weather conditions

Leafminer fly, Liriomyza huidobrensis, under Peruvian lowland and highland conditions

00.020.040.060.080.1

0.120.140.160.18

0 30 60 90 120

150

180

210

240

270

300

330

360

intr

insi

c ra

te o

f in

crea

se

Julian days

11.021.041.061.081.1

1.121.141.161.181.2

0 30 60 90 120

150

180

210

240

270

300

330

360

finite

ra

te o

f in

crea

se

Julian days

0

5

10

15

20

25

30

35

0 30 60 90 120

150

180

210

240

270

300

330

360

net

rep

rod

uct

ion

ra

te

Julian days

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 30 60 90 120

150

180

210

240

270

300

330

360

imm

atu

re s

urv

iva

l

Julian days

0102030405060708090

0 30 60 90 120

150

180

210

240

270

300

330

360

gen

era

tion

tim

e (d

ays

)

Julian days

0102030405060708090

0 30 60 90 120

150

180

210

240

270

300

330

360

Dou

blin

g tim

e (d

ays

)

Julian days

0

5

10

15

20

25

30

35

0 100 200 300

Tem

per

atu

re (

°C)

Julian days-5

0

5

10

15

20

25

0 100 200 300

Tem

per

atu

re (

°C)

Julian days

Lowland agroecology (350 masl)

max.

min.

Lowland highland

Pest management (4)

(Mujica et al.,

in preparation)

Potential establishment (ER) and efficacy (number of generations, GI) of Orgilus

lepidus (Hymenoptera: Braconidae), parasitoid for classical biocontrol of the

potato tuber moth, Phthorimaea operculella

Pest management (4)

ERI GI

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N

nat. increase

r m = 0.1297A

24ºC

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N r m = 0.0622

B18ºC

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 10 20 30 40 50

age (days)

dist

ribut

ion

rate

Egg

Larva

Pupa

Adult

Simulating PhopGV field performance

Pest management (5)

(Sporleder et al. Agriculture and Forest Entomology 2007)

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N

nat. increase

1. Appl.

r m = 0.1297

applications

A24ºC

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N r m = 0.0622

applications

B18ºC

Application of

1.2×1013 granules/hectare

≈≈≈≈ 4,000 LE/hectare

≈≈≈≈ 80% mortality (neonate larvae)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 10 20 30 40 50

age (days)

dist

ribut

ion

rate

Egg

Larva

Pupa

Adult


Pest management (5)


0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N

nat. increase

1. Appl.

2. Appl.

r m = 0.1297

applications 1 - 2

A24ºC

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N r m = 0.0622

applications 1 - 2

B18ºC

Application of




0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 10 20 30 40 50

age (days)

dist

ribut

ion

rate

Egg

Larva

Pupa

Adult


Pest management (5)


0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N

nat. increase

1. Appl.

2. Appl.

3. Appl.

r m = 0.1297

applications 1 - 3

A24ºC

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N r m = 0.0622

applications 1 - 3

B18ºC

Application of




0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 10 20 30 40 50

age (days)

dist

ribut

ion

rate

Egg

Larva

Pupa

Adult


Pest management (5)


0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N

nat. increase

1. Appl.

2. Appl.

3. Appl.

4. Appl.

5. Appl.

6. Appl.

7. Appl.

8. Appl.

r m = 0.1297

applications 1 - 8

A24ºC

0

2

4

6

8

10

12

14

0 30 60 90 120

Time (days)

ln N r m = 0.0622

applications 1 - 8

B18ºC

Application of




0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 10 20 30 40 50

age (days)

dist

ribut

ion

rate

Egg

Larva

Pupa

Adult


Pest management (5)


Global economic assessments/first approximation of crop losses:

Relate pest abundance (pot. no. of generations/season) to % crop loss; consider

precipitation as co-factor

Linking to crop models:

Use of same input data (temperature); daily or hourly simulation interval (e.g.,

using InfoCrop, DSSAT, EPIC).

ILCYM quantify pest populations by separating stage- and age-classes. Individuals

in the damaging age-stage can be quantified over time. These numbers need to be

expressed for a determined leaf area and further calculated into “crop damage-loss

relationships”.

Leaf miner fly: Apply foliar injury (%) crop loss functions in crop models; yields of

early varieties (Desiree) are more affected than of late varieties (Yungay).

y = 1.9501x + 0.1364R² = 0.8941

-10

0

10

20

30

40

50

60

70

0 10 20 30 40

Yie

ld lo

ss (

%)

2008-Desiree

P<.0001

y = 0.6155x - 0.0138R² = 0.7555

-10

0

10

20

30

40

50

0 10 20 30 40 50 60

Yie

ld lo

ss (

%)

Foliar injury (%)

2008-Yungay

P<.0001

Outlook: Integrated assessments

(Muijca, N. & J. Kroschel Crop Protection 2013).Foliar injury (%)

Next steps

ILCYM version 4 with a new interface fully in R-language (shiny-applications) to reduce required skills to maintain ILCYMsoftware and to increase user-friendliness and flexibility to usethe software.

Develop new sub-modules for assessing other influencing co-factors (e.g., precipitation) to improve pest predictions underclimate change on related crop losses.

Combining pest models with crop models in order to betterpredict/quantify losses. Leafminer fly as model pest tointegrate into crop models. Parametrize crop models with cropspecific data at the Peruvian coast.

Acknowledgements

Donors:

• German Federal Ministry for Economic Cooperation

and Development (BMZ/GIZ)

• World Bank

• El Fondo Regional de Tecnologia Agropecuaria (FONTAGRO)

• CRP-RTB

• CCAFS

Thank-you!

next generation crop insect pest models for integrated ... · simon (2013): predicting climate...

Documents