Weather Models and Pest Management Decision Timing

1
Weather Models and Pest Management Decision Timing
Len Coop, Assistant Professor (Senior
Research) Integrated Plant Protection Center,
Botany Plant Pathology Dept. Oregon State
University
2
Topics for today's talk

• Weather data -driven models degree-day and
  disease risk models - concepts and examples
• Some uses and features of the IPPC "Online
  weather data and degree-days" website,
  http//pnwpest.org/wea
• Focus on caneberries and phenology models
• Reasons for modeling

3
Typical IPM questions and representative decision
tools

• "Who?" and "What?"
• Identification keys, diagnostic guides,
  management guides
• "When?"
• Phenology models (crops, insects, weeds), Risk
  models (plant diseases)
• "If?"
• Economic thresholds, crop loss models, sequential
  and binomial sampling plans
• "Where?"
• GPS, GIS, precision agriculture

4
Weather and Degree-day Concepts in IPM

• Degree-days a unit of accumulated heat, used
  to estimate development of insects, fungi,
  plants, and other organisms which depend on
  temperature for growth.
• Calculation of degree-days (one of several
  methods) DDs avg. temperature - threshold. So,
  if the daily max and min are 80 and 60, and the
  threshold is 50, then we accumulate
• (8060)/2 - 50 20 DDs for the day

5
Weather and Degree-day Concepts

• Degree-day models accumulate a daily "heat unit
  index" (DD total) until some event is expected
  (e. g. egg hatch)

Eggs start developing 0 DDs
152
26
126
20
106
22
84
14
Eggs hatch 152 cumulative DDs
70
32
38
cumulative
18
20
20
daily
70o(avg)-50o(threshold)20DD
6
Weather and Degree-day Concepts

• We assume that development rate is linearly
  related to temperature above a low threshold
  temperature

Low temperature threshold 32o F
7
Weather and Degree-day Concepts

• Some DD models sometimes require a local
  "biofix", which is the date of a biological
  monitoring event used to initialize the model
• Local field sampling is required, such as sweep
  net data, pheromone trap catch, etc.

8
IPPC weather data homepage (http//pnwpest.org/wea
)
9
IPPC weather data homepage (http//pnwpest.org/wea
)
10
Degree-day models Examples in pest management

• Nursery crops - Eur. Pine Shoot Moth Begin
  sprays at 10 percent flight activity, predicted
  by 1,712 degree-days above 28 F after Jan. 1st.
• Tree Fruits - Codling moth 1st treatment 250 DD
  days after first consistent flight in traps
  (BIOFIX).
• Vegetables - Sugarbeet root maggot if 40-50
  flies are collected in traps by 360 DD from March
  1 then treat.

11
Degree-day models standardized user interface
12
(No Transcript)
13
Degree-day models Orange tortrix example
14
Degree-day models Orange tortrix example (cont.)
15
Degree-day models forecast weather
16
Thinking in degree-days Predator mites example -
very little activity Oct-Mar (Oct-Apr in C. OR)
Active Period
Active Period
http//pnwpest.org/cgi-bin/ddmodel.pl?sppnfa
17
New version of US Degree-day mapping calculator
1. Specify all regions and each state in 48-state
US 2. Uses all 3200 US weather stations (current
year) 3. Makes maps for current year, last year,
diffs from last year, hist. Avg, diffs from hist.
Avg maps
18
New version of US Degree-day mapping calculator
4. Animated show of steps used to create
degree-day maps
19
5. Revised GRASSLinks interface 6. Improved map
legends
New version of US Degree - day mapping calculator
20
Online Models - IPPC
New - date of event phenology maps we will test
if date prediction maps are easier to use than
degree-day prediction maps
21
Disease risk models

• Like insects, plant pathogens respond to
  temperature in a more-or less linear fashion.
• Unlike insects, we measure development in
  degree-hours rather than degree-days.
• In addition, many plant pathogens also require
  moisture at least to begin an infection cycle.

22
Spotts et al. Pear Scab model (example generic
degree-hour infection risk model) 1.
Degree-hours hourly temperature (oF)
32 (during times of leaf wetness) 2. Substitute
66 if hourly temp 66) 3. If cumul. degree-hours
320 then scab cycle started
23
Some generic disease models applicable to a
variety of diseases and crops Model Disease
Crops
Gubler-Thomas Powdery
Mildew grape, tomato, lettuce, cherry
, hops Broome et al. Botrytis cinerea grape,
strawberry, tomato, flowers Mills
tables scab, powdery apple/pear,
grape mildew TomCast DSV Septoria,
celery, potato, tomato, Alternaria alm
ond Bailey Model Sclerotinia, peanut/bean,
rice, melon rice blast, downy
mildew Xanthocast Xanthomonas walnut --------
--------------------------------------------------
---------
24
Online Models - IPPC
Plant disease models online National Plant
Disease Risk System (in development w/USDA)
GIS user interface
Model outputs shown w/input weather data for
veracity
25
Practical disease forecasts

FIVE DAY DISEASE WEATHER
FORECAST 1537 PDT WED, OCTOBER 01, 2003
THU FRI SAT SUN
MON DATE 10/02 10/03
10/04 10/05 10/06 ...SALINAS PINE... TEMP
74/49 76/47 72/50 72/49
76/49 RH 66/99 54/96
68/99 68/96 58/96 WIND SPEED MAX/MIN (KT)
10/0 10/0 10/0 10/0 10/0 BOTRYTIS INDEX
0.12 0.03 0.09 0.48
0.50 BOTRYTIS RISK MEDIUM LOW
LOW MEDIUM MEDIUM PWDRY MILDEW HOURS
2.0 5.0 6.5 4.0 4.0 TOMATO LATE
BLIGHT READY SPRAY READY READY
SPRAY XANTHOCAST 1 1 1
1 1 WEATHER DRZL
PTCLDY DRZL DRZL DRZL ------------------------
------------------------------------------- TODAY'
S OBSERVED BI (NOON-NOON) -1.11 MAX/MIN SINCE
MIDNIGHT 70/50 ---------------------------------
---------------------------------- ...ALANFOX...FO
X WEATHER...
26
Why weather-driven models for IPM?

• Pest models provide quantitative estimates of
  pest activity and behavior (often hard to
  detect) they can take much of the guess work out
  of timing control measures
• Pest models are expected to become NRCS cost
  share approved practices for certain crops and
  pests, proper spray timing is a recognized
  pesticide risk mitigation practice
• Models can be tied to local biological and
  weather inputs for custom predictions, and
  account for local population variations and
  terrain differences
• Models can be tied to forecasted weather to
  predict future events