|
The apple scab potential module determines the
disease potential for apple scab in the orchard. The goals or disease potential levels are
qualitative. They are: none (no disease potential), low (low disease potential), moderate
(moderate possibility of disease development), high (high possibility of disease
development), severe (disease will develop unless effective steps are taken),
low-post-infection (after infection situation with low disease potential), and high
post-infection (after infection situation with high disease potential). This apple scab
module essentially determines the disease potential of the orchard since the last
pesticide application. The none, low, moderate and high goals (or module end points) rate
the disease potential since the last spray if no new infections have occurred. The low and
high post-infection goals rate the disease potential when there was the possibility of
infection since the last application. For apple scab this determination is based on 5
factors: the current phenology of the orchard, the cultivar susceptibility, the incidence
of scab this season and last and if infections have potentially occurred since the last
fungicide application. The current phenology is determined from orchard profile input
according to a standard classification system illustrated within the system. The cultivar
(from the orchard profile data) susceptibility to apple scab is assigned by the system.
Cultivar disease susceptibility ratings were determined from published information and
through personal communication with growers, orchard consultants and university personnel.
Incidence of scab this season and last season are supplied by the grower within the
orchard profile. The potential for infection since the last fungicide application is
determined by a subroutine. The goals of the infection potential subroutine are: no
infection potential, an infection potential within the last 24 hours, an infection
potential within the last 3 days, an infection potential within the last 4 days and an
infection potential which occurred more than 4 days ago. The factors which determine these
goals include: the amount of rainfall (<2", >2") since the last fungicide
application, application method (both sides (complete) of the tree application or
alternate side of the tree application), days after the last fungicide application that
the infection period ended and infection period calculations based on Mills and Jones et
al. Once the factors which describe the goals of the apple scab module are determined, the
relationship of these factors to each other is decribed.
The specific relationship of the factors which describe a goal is displayed in a
dependency network. Dependency networks are types of decision trees that utilize Boolean
logical approaches. Dependency networks are valuable because they provide a 'hardcopy'
reservoir of system logic and are the means of communication among domain experts and
programmers. The dependency networks are the primary mechanism of communication between
the subject matter experts and the programmers. These 'logic pictures' also provide a hard
copy record of the system's operating logic. This record is most useful in the logic
updating process done by the expert as new knowledge is acquired. The pest management
modules alone represent 135 dependency networks. For example, the 'severe' disease
potential goal is described in a network as two situations which are shown as lines drawn
from the 'and' statements. One situation which describes the 'severe' goal is shown by the
darker lines. The severe disease potential for apple scab can occur from green tip to
second cover phenological stages if the cultivar susceptibility is high and the incidence
of apple scab this season is greater than 0 % and there has been no infection potential
since the last fungicide application. The network describing the severe goal is relatively
simple, involving only two distinct situations. Some networks have many individual
situations to be displayed. The 'High' apple scab disease potential, for example, is
composed of 7 statements which describe 7 different disease scenarios which result in the
"High" disease potential rating.
|
|
The cedar apple rust potential is also
reflected in 3 goals; none, low, and high. The factors which describe each goal are
phenology, cutivar susceptibility, infection period occurrence and historical incidence.
Infection period calculations are based on Aldwinckle et al. In Pennsylvania, cedar apple
rust occurs at 3 levels; it is never a problem, occassionally it is a problem, and
frequently it is a problem. A frequent problem usually indicates that the alternate host
is near-by. A 'high' cedar apple rust potential for an orchard would be given if the
phenology was from pink to petal fall, the cultivar susceptibility was high, there had
been an infection period and the disease was occasionally to frequently a problem in the
orchard. |
