Title: forecasting model for insect pests
1Forecasting Models for major Insect pests
2Forecasting involves all the activities in
ascertaining and notifying the growers of
community that conditions are sufficiently
favourable for certain insect pest, that
application of control measures will result in
economic gain or on the other hand and just as
important that the amount expected is unlikely
to be enough to justify the expenditure of time,
energy and money for control.
Miller and
OBrien (1952)
3Crop Pests - Weather Relationship
Crop
Weather
Pests
4Pre-requisites for developing a Forecast
System
- The crop must be a cash crop(economic yield)
- The insect must have potential to cause
damage(yield losses) - The Insect pest should not be regular
(uncertainty) - Effective and economic control known (options to
growers) - Reliable means of communication with farmers
- Farmer should be adaptive and have purchase power
5National Consultation on a Framework for Climate
Services in Belize
6Criteria for successful Insect pest forecasting
system
- Reliability -use of sound biological and
environmental data - Simplicity - The simpler the system, the more
likely it will be applied and used by producers - Importance -The insect pest is of economic
importance to the crop, - Usefulness -The forecasting model should be
applied when the insect can be detected reliably - Multipurpose applicability -monitoring and
decision-making tools for several diseases and
pests should be available - Cost effectiveness -forecasting system should be
cost affordable relative to available insect pest
management tactics.
7 Experimental pest and disease data
Sensitivity studies
Observed weather
calibration
calibration
Pest and disease models
Weather generator Re-sampling approach
Future hourly weather data
Climate change scenarios
Future pest/disease scenarios
Schematic overview of the Forecast model for
Insect and disease
8(No Transcript)
9Forecast Model- Types
- Between year models
- These models are developed using previous years
data. - The forecast for pests and diseases can be
obtained by substituting the current year data
into a model developed upon the previous years. - Within year models
- Sometimes, past data are not available but the
pests status at different points of time during
the current crop season are available. - In such situations, within years growth model
can be used, provided there are 10-12 data points
between time of first appearance of pests and
maximum or most damaging stage. -
(Amrender,2006)
10Forecast Models Developed in the Past for insect
pests
- Techniques used were essentially Statistical
(Correlation and Regression Analysis) - T.P. Trivedi had proposed a regression model to
predict the pest attack. - Model seems to work only for some years
(1992-1994) - Correlation analysis was used by C.P. Srivastava
to explore the relationship between the rainfall
and pest abundance in different years. - The technique is not effective as the attributes
dont follow normal distribution
11Model based on Regression method
12Weather Related Forecasting Model
- Observations-
- Crop data Phenological development, Growth, Leaf
area and Variety - Insect pest Pest population
- Weather Data Required (hourly for ten or more
- years)
- Precipitation
- Temperature
- Sunshine/cloudiness
- Relative humidity
- Leaf wetness
- Wind direction and speed
National Consultation on a Framework for Climate
Services in Belize
13Degree-Day Models
- Degree-days (DD) are used in models because they
allow a simple way of predicting development of
cold-blooded organisms (insects, mites, bacteria,
fungi, plants). - Degree-day models have long been used as part of
decision support systems to help growers predict
spray timing or when to begin pest scouting. - 1 degree-day (DD)-DD is way of measuring of
Insect growth and development in response to
daily temperature - http//www.ipm.ucdavis.edu/MODELS
14Calculating degree-days
- Degree days (Maximum temperature minimum
temperature)/2 - Base Temperature - Accumulated growing degree days was derived by
using the formula - ????????
- Where,
- Tmax maximum temperature (C)
- Tmin minimum temperature (C)
- Tb base temperature (C)
(Iwata,1984)
15Degree-Days Models types
Simulation model
16Phenology Models
- Phenology models are driven by several weather
parameters on hourly basis. - Relationships between temperature and stage
specific development rates of the insect life
cycles are established in through laboratory
experiments under controlled conditions. - For validation, implemented model predictions are
compared with independent field observations from
several years.
(Stinner et al., 1974)
17Two Methods to Manage Codling moth Larvae
- Calendar Approach
- Treat 3 weeks after full bloom
- Degree Day Model
- Monitor adult flight with pheromone traps
- Biofix 1st consistent catch of moths in traps
- Treat at 250 DD after Biofix
18(No Transcript)
19(No Transcript)
20IPPC weather data homepage (http//pnwpest.org/wea
)
21IPPC weather data homepage (http//pnwpest.org/wea
)
22GDD approach
- This method is based on the assumption that the
pest becomes inactive below a certain temperature
known as base temperature - GDD ? (mean temperature base temperature)
- Not much work on base temperature for various
diseases and insect pest. Normally base
temperature is taken as 50 C - Under Indian conditions, mean temperature is
seldom below 50 C - Need for work on base temperature and initial
time of calculation
23(No Transcript)
24Limitations of Degree-Day Models
- Insect response to temperature is not linear
- Lower Thresholds Temperature known for very few
species. - Measured temperatures not the same as those
experienced by the pest.
25Models developed at IASRI
- Mustard
- Aphid
- Cotton
- American boll worm
- Pink boll worm
- Spotted boll worm
- Whitefly
- Groundnut
- Spodoptera litura
- Onion
- Thrips
- Sugarcane
- Pyrilla
- Early shoot borer
- Top borer
- Pigeon pea
- Pod fly
- Pod borer
- Rice
- BPH
- Gall midge
- Mango
- hoppers
- fruit-fly
26Forecasting model for H. armigera
S. No. Parameters Pest infestation in 10-14 standard week Pest infestation in 10-14 standard week Pest infestation in 10-14 standard week
S. No. Parameters High Medium Low
1 Sudden rise in the minimum temperature by gt50C around 7-8 standard week - - -
2 Rainfall during 1-9 standard weeks - - -
3 Base adult moth population gt15 per week during 5-7 standard week - - -
(Indian Institute of Pulses Research)
27Forewarning model given by National research
Center for Rapeseed Mustard
28(No Transcript)
29- Potato aphid
- Potato aphid (Myzus persicae) is an abundant
potato pest and vector of potato leaf-roll virus,
potato virus Y , etc. - Potato aphid population Pantnagar (weekly
models) - Data used 1974-96 on MAXT, MINT and RH
- X1 to X3) lagged by 2 weeks
- Model for December 3rd week
- Y 80.25 40.25 cos (2.70 X12 - 14.82)
- 35.78 cos (6.81 X22 8.03)
-
Trivedi et al 1999
30Potato Aphid population in 3rd week of
December at Pantnagar
Trivedi et al 1999
31(Yadav and singh,2015)
32Deviation method
- Useful when only 5-6 year data available for
different periods - Week-wise data not adequate for modeling
- Combined model considering complete data.
- Not used for disease forewarning but in pest
forewarning - Assumption pest population in particular year
at a given point of time composed of two
components. - Natural cycle of pest
- Weather fluctuations
(Mehta et al,2001)
33- Mango
- Mango fruit fly Lucknow (weekly models)
- Data used 1993-94 to 1998-99 on MAXT, MINT and
RH X1 to X3 - Model for natural pattern
t Week no. Yt Fruit fly population count
at week t
Mehta et al.(2001)
34Mango fruit fly population prediction model based
on deviation method in Lucknow
Mehta et al.(2001)
35Pest simulation models
- Pre requisites for Simulation models
- Mathematical descriptions of biological data.
- Computer programs or software to run these
models. - Application of these models in understanding
population dynamics and dissemination of pest
forecasts for timely pest management decisions. - (Coulson and Saunders, 1987 )
36EPIPRE
- EPIPRE (EPidemics PREdiction and PREvention) is a
system of supervised control of diseases and
pests in wheat. - The participating farmers do their own pest
monitoring, simple and reliable observation and
sampling techniques. - Farmers send their field observations to the
central team, which enters them in the data bank.
Field data are updated daily by means of
simplified simulation models. Expected damage and
loss are calculated and used in a decision
system, that leads to one of three major
decisions - treat
- don't treat
- make another field observation
- Ex-Rhopalosiphum padi
37Organization scheme for EPIPRE in relation to
farmers and research organizations
38Generic Pest Forecast System (GPFS)
- Combination of multiple weather variables and
biological processes into a single predictive
model Temperature, relative humidity, leaf
wetness (hourly) - Growth, mortality, infection
- Predict timing and abundance of specific life
stages - Estimate damage to specific host plants based on
pest and host phenology
Hong et al,2013
39GPFS Model Modules
- Development Response to temperature
- Simple linear model (Tmin, Topt1, Topt2, Tmax)
- Mortality factors Heat, cold, aging, soil
moisture, food availability
- Case Studies
- Insects
- Bactrocera dorsalis (oriental fruit fly)
- Epiphyas postvittana (light brown apple moth)
- Helicoverpa armigera (boll worm)
Hong et al,2013
40Oriental Fruit Fly
- Highly polyphagous and extremely invasive
- Development is temperature driven
- Available data sets of distribution records and
seasonal observations available for validation - Model validation in three locations India, USA
(HI), China - Comparisons made with CLIMEX-compare locations
41Case Study I Bangalore, India
Of fruit fly
(Jayanthi and Verghese, 2011)
42Ordinal logistic model model for qualitative
data
- pest / disease outbreak can be taken even if the
information on the extent of severity is not
available but merely the epidemic status is
accessible - models have added advantage that these could be
obtained even if the detailed and exact
information on pest count / disease severity is
not available but only the qualitative status. - where z is a function of weather variables.
- Forecast / Prediction rule
- If P .5 more chance of occurrence of epidemic
- If P lt .5 probability of occurrence of epidemic
is minimum
(Mehta et al. 2001 Mishra et al. 2004 Johnson
et al. 1996 Agrawal, et al. 2004)
43Forecasting outbreak of pest using Ordinal
Logistic model
Crop (Location) Pest Important variables Time of Time of Workers
Damage Forecast
Cotton (south India) Whitefly MAXT,MINT,RH I RH II Mid Dec. Mid Nov. Agrawal et al. (2004)
Sugarcane (Muzaffarnagar) Pyrilla MAXT RHM Oct.-Nov. May Mehta et al. (2001)
Mango (Lucknow) Fruit fly MAXT,MINT RH I May-June 2nd week of March Misra et al. (2004)
MAXT maximum temperature, MINT minimum
temperature, RH 1 relative humidity
(morning),RHIIrelative humidity (evening) and
RHMmean relative humidity
44Uses of ANN(artificial neural network technique)
- ANN provides an efficient alternative tool for
forecasting. - Neural Networks dont make any distributional
assumption about the data. - It learns the patterns in the data, while
statistical techniques try to do model fitting. - ANNs can often correctly infer the unseen part
of a population even if data contains noisy
information. - This makes neural network modeling a powerful
tool for exploring complex, nonlinear biological
problems like pest incidence.
(Agrawal et al. 2004 Dewolf et al. 1997, 2000
Kumar, et al. 2010)
45Remote Sensing
- Remote sensing technologies are used to gather
information about the surface of the earth from a
distant platform, usually a satellite or airborne
sensor - The integration of remote sensing, GPS and GIS
are valuable tools that can enable resource
managers to develop maps showing the distribution
of insect infestations over large areas - Site-specific data, such as type of insect, level
of damage, stage of attack, yield loss are
collected from different sources, stored and
managed in spatial database , either contained
within the GIS or connected to the GIS from an
external source.
(Wu et al., 2008)
46Predicting the potential geographical
distribution of sugarcane wooly aphid
- Prediction of potential geographical distribution
of sugarcane wooly aphid using its current
distribution and data on range of environmental
parameters. - Two approaches for the purpose
- GPS (Global Positioning System)
- GIS (Geographic Information System)
(Prabhakar et al., 2012)
47Infrastructure to calculate risk maps
step1
met. data
Geo.data
step 2
combine with GIS
Interpolation
step 3
Calculation of forecasting models with
interpolated input parameters
step 4
Presentation of results
48The forecasting models database
- The CIPRA (Computer Centre for Agricultural Pest
Forecasting) software allows the user to
visualise forecasts of insect development . - Within CIPRA, there is forecasting models for a
total of 35 pests (25 insects and 10 diseases).
In vegetables ( cabbage, carrot, onion, potato,
tomato), fruits(apple, grape, strawberry), and
cereals (wheat, barley, corn). - Each crop has its own independent computer
programs file, known as DLL (Dynamic Link
Library), which makes possible their integration
in other specialized software. - (Bourgeois et al., 2008)
49DEGREE-DAYS MODELS AVAILABLE IN CIPRA
- Mathematical models vary from simple degree-days
approach based on air temperature to more
detailed epidemiological system based on air
temperature, relative humidity and duration of
leaf wetness.
- Crop
- Apple
- Crucifers
- Potato
- Tomato
- Cotton
- Insect
- Codling moth
- Diamond back moth
- Colorado potato beetle
- Tomato fruit borer
- Spotted boll worm
50(No Transcript)
51(No Transcript)
52Computer Software Model for Prediction for Insect
pest
- PEST-MAN is a computerized forecasting tool for
apple and pear pests-Canada - MORPH is predictive computer model for
horticultural pest-UK - SOPRA is applied as a decision support system for
eight major insect pests of fruit orchards
-Switzerland and southern Germany - The SIMLEP decision support system for Colorado
potato beetle (Leptinotarsa decemlineata)-Germany
and Austria
53 mainstream technology of computer software for
forecasting
54(No Transcript)
55Insect pest forecasting concept
56(No Transcript)
57(No Transcript)
58(No Transcript)
59success of a forecasting
- The success of a forecasting system depends,
among other things, on - The commonness of epidemics (or need to
intervene) - The accuracy of predictions of epidemic risk
(based on weather, for example) - The ability to deliver predictions in a timely
fashion - The ability to implement a control tactic
(Insecticide application, for example) - The economic impact of using a predictive system
60Uses of insect pest forecasts
- Forewarning or assessment of insect pest
important for crop production management - for timely plant protection measures
- information whether the insect pest status is
expected to be below or above the threshold level
is enough, models based on qualitative data can
be used qualitative models - loss assessment
- forewarning actual intensity is required -
quantitative model - For making strategic decision-
- Prediction of the risks involved in planting a
certain crop. - Deciding about the need to apply strategic
control measures (soil treatment,
planting a resistant cultivar etc)
61- For making tactical decision-
- Deciding about the need to implement insect pest
management measure - Entomologist and meteorologists have often
collaborated to develop insect pest forecasting
or warning systems that attempt to help growers
make economic decisions for managing insect. - These types of warning systems may consist of
supporting a producers decision making process
for determining cost and benefits for applying
pesticides, selecting seed or propagation
materials, or whether to plant a crop in a
particular area. - forecasting models in order to optimize timing of
monitoring, management and control measures of
insect pests
62Limitations of forecasting Models-
- Statistical models are simple in their usage and
less parameter-intensive, but they are limited in
the information they can provide outside the
range of values for which the model is
parameterized. - Crop simulation models are rather hard to use and
parameterize. The need for calibration can be
quite data extensive and not applicable to some
developing countries. - The major limitation in use of satellite-borne(rem
ote Sensing) data in pest forewarning is the
timely availability of cloud-free data with the
desired spatial and spectral resolution
63Thank You