Luis del Río North Dakota State University Developing a Sclerotinia stem rot warning system for canola in North Dakota.

Luis del RíoNorth Dakota State University

Developing a Sclerotinia stem rot warning system for canola in

North Dakota

Photo: www.canolacouncil.org

Photo: www.canolacouncil.org

North Dakota canola producing areas

Divide Burke Renville Rolette Towner Cavalier

Mountrail

Williams

Ward

McHenry Pierce

Benson

Ramsey

Wells Foster

Eddy

Stutsman

GriggsMcLean

Bottineau RosseauKittson

Marshall

PenningtonWarren

Red Lake

Sheridan

Nelson

Fargo

Sclerotinia stem rot disease cycle (Sclerotinia sclerotiorum)

A. Lamey

Model development

206 fields from 25 locations per year

SSR incidence on 50 plants per field

Annual field survey (third wk August)

Disease data collection

SSR mean incidence per location/ year

NDAWN stations Hourly data on air and soil temp, solar

radiation, dew temp, wind speed, etc. Means and new variables created by

periods of 15 days

Weather data collection

Model development

Logistic regression analysis Model produced using 177 data points

collected between 2001 and 2006 Model validation using cross-validation

option of SAS

Model development and validation

Model development

Risk of SSR development calculated every 3 days

Color maps available to growers through web site during canola flowering period

Model forecasting

Model development

Incidence of Sclerotinia stem rot of canola in North Dakota between 1995 and 2008

'95

'96

'97

'98

'99

'00

'01

'02

'03

'04

'05

'06

'07

'08

02468

1012141618

Inci

denc

e (%

)

Years

Results

Most important weather variables were precipitation and solar radiation

Three models developed, one for second half of June, one for first half of July and a general model

Results

Percent concordance 77 80 83

Percent discordance 23 20 17

Somers’ D 0.54 0.59 0.66

Gamma 0.54 0.59 0.66

Parameters Model 1 Model 2 Model 3

Tau- a 0.27 0.30 0.33c 0.77 0.79 0.83

Resultsy= 4.08+0.05(rain3)+0.07(rain4)-0.55(solar4)+0.53(rain5)

-0.20(solar5)+0.38(rain6)

Results

Risk estimation North Dakota, 2007

Flowering period

Low riskIntermediate riskHigh risk

July 21

July 18

July 15

July 12

July 9

July 6

July 3

June 30

June 27

June 24Cavalier Towner Rolette

3% 8%4%Mean incidence

Results

WardBensonRosseauBottineauBensonTownerCavalierRollette

33 44 0100 44 53 76100

2 1 025 7 3 4 8

Prevalence (%) Incidence (%)County

SSR county prevalence and incidence, ND 2007

Results

CavalierRamseyRoseauBensonBottineauRollaTownerWard

68 78100 80 7100 28 5

5 414 2 1 7 1 1

Prevalence (%)

Incidence(%)County (%)

SSR county prevalence and incidence in 2008

47 9 5 515 91820

Fields

Results

WardBensonBottineauTownerCavalier

13 25 50 60 47

1 2 1 2 3

Prevalence (%) Incidence (%)County

SSR county prevalence and incidence, ND 2009

Results

SS

R P

reva

len

ce (

%)

Number of high-risk warnings0 1 2 3 4 5 6 7 8 9 10

0

20

40

60

80

100

Association between number of high-risk warnings and SSR prevalence

Results

Summary

Empirical model developed using weather data

Overall accuracy of model is approximately 75%

Prevalence of high risk warnings associated with higher SSR prevalence

Model improvement is needed Areas were collaboration is sought:

- Modeling leaf wetness duration- Use of Doppler radar data- Quantification of risk of apothecia

formation

Summary

Acknowledgements

Curt Doetkott, ITS NDSU scouts and county

agents USDA-ARS/Sclerotinia

Initiative North Dakota Canola

Growers Association