Northeast Region IPM Grants
|
|
States: New York, West Virginia, Investigators: Turechek,W.W., A.R. Biggs, G.W. Lightner, H.S. Aldwinckle , D.I. Breth, D.M. Glenn Institutions: Cornell University,
West Virginia UniversityProject Type: research and extension Award*: $185,177 Term: 36 months beginning 7/1/01 Crop:
apples and pears
*Award shown is total amount to be used over the course of the project term.
Fire blight, caused by the bacterium Erwinia amylovora, is one of the most
destructive and difficult-to-control diseases of apple. Over the last decade,
major changes in horticultural practices have increased the chances for infection
and level of damage likely to occur. The most widely used approach for disease
management in the Northeast is to apply a series of streptomycin antibiotic
sprays during the blossoming period; 2-4 routine treatments per year in most
locations, but some use 8 or more. This approach seldom affords complete control,
sometimes fails and, because outbreaks are so erratic, often results in unnecessary
treatments when conditions do not support infection. While many growers are
willing to forego the cost of possibly ineffective treatments as insurance against
the potential losses of fire blight, unnecessary applications are costly. Aside
from the added costs, excessive use of antibiotics can lead to the appearance
of resistant strains in the pathogen population and, quite possibly, in untargeted
bacterial populations.
The ability to predict the onset of fire blight has been the most limiting factor
in improving the overall management of the disease. MARYBLYT is a computer program
for forecasting fire blight that predicts the four distinct types of infection
events (i.e., blossom, shoot, canker, and trauma blight) and the appearance
of symptoms that follow. MARYBLYT was first introduced in 1989, released for
commercial distribution in 1992, and is now used by fruit growers in research,
teaching and extension programs in 30 US states and over 20 countries worldwide.
In the blossom blight submodel of MARYBLYT, the minimum conditions necessary
for blossom infection are: 1) flowers open with stigmas and petals intact; 2)
accumulation of at least 110 cumulative degree hours greater than 18.3 C from
the start of bloom; 3) a wetting event of at least 0.25 mm of rain or heavy
dew or a rain of 2.5 mm or more the previous day; and 4) an average daily temperature
of 15.6 C. MARYBLYT characterizes risk as either low, moderate, high or infection
depending on whether one, two, three, or four of the risk factors have exceeded
their minimum values. The faults with this approach are: 1) MARYBLYT assumes
each risk factor contributes equally to the risk of developing blossom blight;
2) the risk of infection does not increase relative to the degree in which individual
factors exceed their minimums; 3) MARYBLYT does not account for varietal susceptibility,
orchard age, or inoculum pressure (factors that may dramatically reduce (or
increase) the risk of infection); and 4) the interpretation of the output is
subjective. As a result, MARYBLYT tends to predict infections when none occur;
resulting in the needless applications of antibiotics.
We propose to: 1) Revise MARYBLYT to calculate a system of 'risk points' as
function of the MARYBLYT risk factors, then using the revisions, evaluate various
management-action thresholds based on the accumulation of 'risk points'; 2)
Modify MARYBLYT to account for varietal susceptibility, orchard age, and inoculum
potential; and 3) Develop a user-friendly, windows-based version of MARYBLYT.
The revised model will be statistically evaluated using receiver operating curve
analysis and field validated in unsprayed test blocks. The modifications will
incorporate flexibility into MARYBLYT; allowing users to choose suitable action
thresholds based on variety, orchard age, inoculum pressure, and their comfort
for assuming risk. Ultimately, these changes will reduce the cost of managing
fire blight by emphasizing efficient, and minimizing unnecessary, applications
of antibiotics.
Fire blight, caused by the bacterium Erwinia amylovora, is one
of the most destructive and difficult-to-control diseases of apple and pears
(1,36,41). Throughout the nearly 200 year history of fire blight the disease
has been an elusive malady in that sometimes severe epidemics develop in young
orchards with no history of the disease and sometimes few symptoms appear in
established orchards with a recent history of severe blight. Epidemics develop
quickly, destroying blossoms, vegetative shoots, major limbs and, sometimes,
whole trees. The 2000 growing season was perhaps one of the worst years for
fire blight ever recorded in the Midwest and Northeast United States (5,17,29).
In fact, that the US House of Representative's Agriculture Appropriations Conference
Committee has appropriated $38 million in the fiscal 2001 agriculture spending
bill (HR 4461) for losses incurred as a partial result of fire blight. It is
years like these that emphasize the need to improve our understanding of fire
blight so that we better develop our tools for disease management, especially
in the face of a depressed and rapidly changing global market where a single
unprofitable year can force a grower out of business (44).
Over the last decade, consumer and market demands have forced major changes
in horticultural practices. These changes have not only increased chances for
infection but the level of damage/infection likely to occur (41). For example,
high density orchards of 250-500 trees/acre are replacing older orchards with
80 to 120 trees/acre. Clonal apple rootstocks with uniform susceptibility to
infection (e.g., M.9, M.26) are almost exclusively used in newly-planted orchards.
It is now known that systemic invasion of rootstocks resulting from blossom
infection is common and, in a large proportion of instances, this invasion can
cause girdling that can kill a tree within one season (26,41). The choice of
rootstocks and management practices encourage early bearing (second to third
leaf). Perhaps most important, is the widespread increase in new high market
value apple cultivars like Gala, Fuji, Honeycrisp, Jonagold, Braeburn and others
that are highly susceptible to fire blight (41).
Most contemporary control programs emphasize thorough orchard sanitation and
dormant pruning to remove or reduce sources of inoculum, early season applications
of copper materials, limited use of nitrogen to avoid an excess of succulent
growth, insect control, and the frequent use of protective bactericide's through
the highly susceptible flowering period every year to prevent primary infections
(4). The most widely used approach in the U.S. is to apply a series of streptomycin
antibiotic sprays at frequent intervals during the blossoming period (4,20);
2-4 routine treatments per year in most locations, but some use 8 or more when
shoot blight is also targeted. This approach, while generally adequate, seldom
affords complete control, sometimes fails and, because outbreaks are so erratic,
often results in unnecessary treatments when the conditions do not support infection.
While many growers are willing to forego the cost of possibly ineffective treatments
as 'insurance' against the potential losses of fire blight, unnecessary applications
are costly and the excessive use of antibiotics can lead to the appearance of
resistant strains of the pathogen (18,19,20,41). Furthermore, the use of antibiotics
as a means of pest control is problematic and is scrutinized by both the public
(15) and scientific community (20).
Until recently, our ability to predict the onset of fire blight epidemics accurately
and reliably has been the most limiting factor in improving the overall management
of the disease. MARYBLYTTM (16,35) is a computer program for forecasting fire
blight in apples and pears that predicts the four distinct types of infection
events (i.e., blossom, shoot, canker, and trauma blight) incited by E. amylovora
as well as the appearance of symptoms that follow (33,34). It was first introduced
in 1989, released for commercial distribution in 1992 (16,35), and is now used
by fruit growers and in various research, teaching and extension programs in
30 US states and over 20 countries worldwide. The program's popularity and widespread
use are attributed to: 1) The destructive nature of fire blight and the high
costs of control; 2) Its ability to predict specific infection events far enough
in advance that protective treatments can be made and eradication measures can
be timed for maximum effectiveness; 3) Data inputs are relatively simple and
easy to acquire; 4) The program claims insensitivity to geographical climate
differences, operates independently from calendar dates, and can be used with
either U.S. or metric units; and 5) Predictions are obtained within minutes
and are accompanied by a variety of visual and audio prompts are given with
respect to risks and treatment warnings.
Blossom blight is the most threatening and destructive phase of the disease;
providing the inoculum for the shoot, root, and trauma blight phases (1,4,33).
As a result, management practices focus on controlling this phase. In the blossom
blight submodel of MARYBLYT four risk factors are monitored to identify possible
infection events. The risk factors and the associated minimum conditions necessary
for blossom infection are as follows: 1) flowers open with stigmas and petals
intact; 2) accumulation of at least 110 cumulative degree hours (CDH) greater
than 18.3 C from the start of bloom; 3) a wetting event of at least 0.25 mm
of rain or heavy dew or a rain of 2.5 mm or more the previous day; and 4) an
average daily temperature of 15.6 C (33,41). MARYBLYT characterizes risk as
either low, moderate, high, or infection depending on whether one, two, three,
or all four of the risk factors have exceeded their minimum values. Observation
has shown that when all 4 of these parameters were met, early symptoms of blossom
infection could be predicted and observed with the accumulation of 57 degree
days >12.8 C after an identified infection event. It was also found that
the first symptoms of shoot blight occurred approximately 57 degree days >12.8oC
after the appearance of either blossom blight symptoms, except in years when
the appearance of the winged adults of the white apple leafhopper was delayed
(33).
Despite its appeal, MARYBLYT is not a perfect forecaster. The model tends to
predict infections when none occur, especially in less susceptible varieties
and in areas with no history of fire blight; resulting in the needless applications
of antibiotics (13,32,42). The excessive use of antibiotics promotes the development
of antibiotic resistant strains of the pathogen and, potentially, in untargeted
populations of bacteria (15,20). As output, the model only generates qualitative
assessments of infection potential without regard to which risk factors have
exceeded their minimums or, for those factors that have exceeded their minimums,
to what degree have the minimums been exceeded. As Steiner observed: "When
environmental conditions meet these criteria only marginally, the incidence
of blossom blight infections is usually low with severity varying due to individual
site differences (variation in bloom, elevation, local dews, blight history,
grower management practices, etc.). By contrast, severe epidemics affecting
large areas are most likely to occur when all or several of the criteria are
well above the minimum activity thresholds" (33). This is particularly
problematic when a grower attempts to factor in the influence of varietal resistance,
orchard age, or inoculum pressure. These factors are known to play a role in
fire blight susceptibility but are not taken into account by MARYBLYT (1,41).
A newer fire blight forecaster, named Courgarblight, was developed by Tim Smith
at Washington State University (32). The model was developed in response to
the poor performance of other fire blight risk assessment models when used in
the Pacific Northwest; MARYBLYT tended to over-predict infections. The inaccuracies
of MARYBLYT were partially attributed to how MARYBLYT factors in 'average temperature'
and its lack of consideration of inoculum pressure. Using three years of weather
data, Breth et al. (4) compared the performance of MARYBLYT and Cougarblight
under New York conditions and found that they performed similarly. Cougarblight
shares some of the same shortcomings of MARYBLYT, for example varietal susceptibility
is not explicitly defined in the model. However, Cougarblight is not (yet?)
a stand-alone program like MARYBLYT; predictions are obtained using a 'lookup'
chart. In truth, either of these two models could be targeted for improvement
based on the limitations outlined below, but we chose MARYBLYT because of its
wide usage and acceptability in the Northeast and it is ready availability as
a stand-alone program.
Justification
Despite its wide-scale use, major limitations to MARYBLYT are: (1) objective
aspects are limited to qualitative predictions (i.e., + an infection event);
quantitative assessments depend on experience and subjective considerations
(i.e, degree to which minimum thresholds are exceeded, rate and timing of epiphytic
inoculum potential increase; level of previous fire blight and thoroughness
of blight management program); (2) MARYBLYT assumes that each risk factor contributes
equally to the risk of developing blossom blight; (3) the risk of infection
does not increase relative to the degree in which individual risk factors exceed
their minimums; (4) MARYBLYT assumes that risk the factors are uncorrelated
i.e., the cumulative effect of the factors on blossom blight are considered
neither synergistic nor antagonistic; (5) MARYBLYT lacks varietal specificity;
i.e., all varieties are considered equally susceptible; (6) the model is based
on the assumption of abundant inoculum; and (7) growers and extension agents
often find the current DOS-based version of MARYBLYT out-of-date and difficult
to use. Because of these limitations, growers sometimes make the decision to
spray when it is not needed (e.g., because varietal resistance compensated for
exceeding the MARYBLYT high risk advisory) resulting in the inefficient and/or
excessive use of antibiotics and potentially fostering resistance development.
Or growers failed to spray when needed resulting in economic losses to the grower
(e.g., because the grower could not gauge the risk of infection, for example,
when two of the MARYBLYT criteria greatly exceeded their minimums).
Other successful forecasters, such as those developed for powdery mildew of
grapes caused by Uncinula necator (8), sclerotinia stem rot in oil seed rape
caused by Sclerotinia sclerotiorum (39), alternaria leaf blight of carrots caused
by Alternaria dauci (7), late blight of potato caused by Phytophthora infestans
(14), downy mildew of hops caused by Pseudoperonospora humuli (30), and apple
scab caused by Venturia inaequalis (23) are based on the accumulation of 'risk
points' (typically as a function of weather and important crop factors). When
the summation of risk points exceed some target level (i.e., a threshold) then
a management action is taken. The advantage of such an approach is that disease
pressure is directly related to the accumulation of risk points. This incorporates
flexibility in the model because action thresholds (based on the accumulation
of risk points) can vary in accordance to a grower's comfort level for assuming
risk, market fluctuations, varietal differences (if varietal susceptibility
is not explicitly defined in the forecaster), etc.
Inherent to the success of these forecasters is the choice of appropriate risk
factors. The risk factors chosen for the MARYBLYT blossom blight submodel came
about as the result of extensive laboratory investigations and empirical field
evidence (3,24,28,37,38,43,45,46). Thus, the set of risk factors is not in question,
but rather the relationship among factors and their defined minimums relative
to disease development, especially on different varieties. As apple production
changes, so does the risk of fire blight. Some practices increase the risk,
some may decrease it. MARYBLYT must be flexible in order accommodate these changes
if we intend to use this model in the future. By evidence of its current widespread
usage, MARYBLYT has the potential to greatly impact how we manage this disease.
Failed predictions or numerous over-predictions will cause growers to abandon
MARYBLYT. However, if we address the weaknesses now, MARYBLYT's use in the industry
will expand, allowing efficient and minimal use of antibiotics for fire blight
management.
1) Revise MARYBLYT to generate a system of 'risk points' based on a quantitative measurement of risk of infection (i.e., a probability). Using the system of 'risk points', determine optimal threshold values for minimizing blossom blight.
2) Modify MARYBLYT to account for varietal susceptibility and inoculum pressure.
3) Develop a user-friendly, windows-based version of MARYBLYT.
|
2001 Northeast IPM Funded Projects |
Home |
What Is It?: |
|