Adoption of Integrated Pest Management by apple growers: the role of context

Geo Kaine
a
and Denise Bewsell
b
*
a
Department of Primary Industries, Tatura, Tatura, Victoria, Australia;
b
AgResearch Ltd, Lincoln Research Centre,
Christchurch, New Zealand
(Received 2 October 2007; nal version received 18 March 2008)
The adoption of pest and disease management practices has been the subject of numerous studies. Unfortunately,
there is little consistency in their ndings with regard to the variables that inuence growers decisions to adopt
these techniques. In this study we focus on context as a means of explaining the lack of consistency. We used the
results of in-depth interviews and a mail survey to explore context as a predictor of Australian apple growers
behaviour with respect to the management of codling moth and pest mites. We found that climate, topography,
spatial separation between orchards and the crop mix determined the type and intensity of pest and disease
pressures experienced by growers. Given the types and intensities of pest and disease pressures present, the
management practices growers used depended on the range and eectiveness of the control options that were
available to them. We concluded that variables representing specic aspects of orchard context were the primary
determinants of the particular combinations of pest and disease management practices used by apple growers, not
variables representing demographic and general enterprise characteristics that have been proposed in past studies.
Keywords: adoption; Integrated Pest Management (IPM); apple growers; context; Australia
1. Introduction
The adoption of pest and disease management
practices, especially Integrated Pest Management
(IPM), has been the subject of numerous studies.
The latter have investigated the adoption of practices
such as monitoring of pests and disease, the use of
population thresholds to determine spray regimes,
monitoring and use of benecial insects to control
pests, the use of selective chemicals, the use of growth
regulators and the use of mating disruption techni-
ques. Unfortunately, there is little consistency in their
ndings.
We present here the results of a study into the role
of context in adoption of IPM among apple growers
in Australia in an attempt to explain the inconsis-
tencies in the ndings among previous studies.
Previous work (Kaine et al. 2005; Boland et al.
2006) has indicated that context is important when
investigating the adoption of technologies by primary
producers. We focus specically on two apple pests
codling moth and pest mites and the adoption of the
IPM techniques associated with controlling them.
2. Past studies
In many studies on the adoption of IPM techniques
researchers have concentrated on identifying relation-
ships between the adoption of pest and disease
practices and the characteristics of growers and their
enterprises. Higher levels of formal education or
schooling in less developed countries have been
positively associated with the adoption of IPM
practices such as monitoring of pest levels, use of
selective chemicals and encouragement of benecial
insects in a variety of industries (Ridgley and Brush
1992; Rama Rao et al. 1997; Waller et al. 1998;
Chaves and Riley 2001; Shennan et al. 2001).
However, Grieshop et al. (1988) found no relation-
ship between education and adoption of IPM among
tomato growers in the United States, while Chaves
and Riley (2001) found that the signicance of the
role of education varied among coee growers in
Colombia, depending on the mix of practices they
analysed. Brough, Frank, Page, and Lindsay (1996)
recorded a negative relationship between experience
and adoption of IPM for mites.
Studies into the relationship between choice of
pest and disease management techniques and farm
size have also yielded inconsistent ndings. Shennan
et al. (2001) recorded a correlation between choice of
pest and disease management techniques and farm
size for vegetable and fruit growers in California.
Fernandez-Cornejo (1998) obtained a similar result
when investigating IPM amongst viticulturists. How-
ever, Chaves and Riley (2001), in their analysis of pest
management practices among Colombian coee
*Corresponding author. Email: denise.bewsell@agresearch.co.nz
International Journal of Pest Management
Vol. 54, No. 3, JulySeptember 2008, 255265
ISSN 0967-0874 print/ISSN 1366-5863 online
2008 Taylor & Francis
DOI: 10.1080/09670870802065256
http://www.informaworld.com
growers, found a relationship between farm size and
adoption for some combinations of practices but not
for other combinations of practices. In contrast,
Grieshop et al. (1988), Ridgley and Brush (1992) and
Waller et al. (1998) found no correlation between
farm size and adoption of IPM among tomato, pear,
or potato growers, respectively, in the United States.
Finally, Fernandez-Cornejo et al. (1994) obtained
mixed results when investigating the impact of risk
aversion on the adoption of IPM among vegetable
growers in the United States.
Several studies have described the complexity of
pest and disease management (for example, Grieshop
et al. 1988; Rodr guez and Niemeyer 2005). Research-
ers have suggested that nancial incentives could help
promote IPM use, particularly in the transition from
conventional management to integrated management
of pests and diseases (Urquhart 1999; Brewer et al.
2004).
In several studies researchers have assumed the
failure to adopt IPM to reect a lack of skill, and they
have focused on the need to educate farmers.
Generally this has centred on helping farmers to
recognise pest or disease symptoms (van Mele et al.
2001; Nyankanga et al. 2004; Poubom et al. 2005).
Other studies have investigated the inuence of the
media and extension agents on growers decisions
(van Mele et al. 2002; Malone et al. 2004).
In related studies, some researchers have attrib-
uted the rapid spread of IPM techniques, particularly
in Asian and African countries, to novel approaches
to grower education and extension (Heong and
Escalada 1998; Norton et al. 1999). Escalada and
Heong (1993) attributed the slow spread of IPM
techniques among rice farmers in the Philippines to a
lack of knowledge among growers, and concluded
that farmer eld schools would accelerate adoption,
by providing growers with the opportunity for
experiential learning of IPM skills.
Both Jeger (2000) and Kogan (1998) concluded
that the success of the farm eld schools in Southeast
Asia had little to do with a new or novel approach to
extension. Rather, the apparent achievements of
farmer eld schools in promoting changes in the
management of rice pests were due to the banning of
widely used broad spectrum insecticides. For exam-
ple, in Indonesia, 57 broad spectrum organopho-
sphate, pyrethroid and chlorinated hydrocarbon
insecticides were banned by presidential decree
(Kogan 1998). The only option for rice growers in
these circumstances was to adopt IPM strategies,
which meant gaining an understanding of the
pest and disease cycles by attending farmer eld
schools.
Jeger (2000) also cites the case of wheat produc-
tion in the United Kingdom, which has been free of
pest or disease crises. He argues that there is a low
level of awareness of IPM amongst wheat producers
because current approaches to pest and disease
management remain successful. Consequently, there
is no incentive to develop IPM strategies for this
industry.
Jeger (2000) went on to explore the dierences
between rice and vegetable production in Asia, and
to comment on the potential for farmer eld schools
to promote IPM in vegetables. He argued that the
potential is limited at present, especially as the
industry is dominated by non-indigenous crops.
Currently, vegetable crops are grown in biologically
diverse home gardens, which provide a buer against
pests and diseases (Jeger 2000).
We found there was little consistency in the
ndings of studies into the adoption of pest and
disease management practices. Studies focusing on
grower and enterprise characteristics have failed to
identify any relationships between variables such as
age, education and experience that are consistent
across industries and countries. This accords with the
observation of Abadi Ghadim and Pannell (1999)
that overall, despite numerous studies, the results of
research in this eld [adoption of innovations] have
been disappointing . . . the results of dierent studies
are often contradictory regarding the importance and
inuence of any given variable (p. 145).
3. The impact of context on adoption
Very few studies have focused on the inuence of
context on the pest and disease management decisions
of growers. Williamson et al. (2003, p. 196) observed
that the way farmers perceive the benets and costs
of dierent pest management strategies will have an
impact on their pest management decisions. Orr and
Jere (1999) used cluster analysis to determine target
groups for IPM amongst smallholders, i.e. farmers
with small areas of land and crop, in Malawi. They
clustered their sample on four variables; the gender of
the head of the household, whether the household
was self-sucient in maize, cash crops grown and
farm size. Five distinct groups resulted and they
developed recommendations outlining a mix of IPM
strategies for each (Orr and Jere 1999).
Here, we drew on Kaines (2004) framework for
identifying the contextual factors that inuence the
adoption of farming practices, in order to clarify the
importance of contextual factors in the adoption of
IPM relative to demographic and other factors. In
doing so, we sought to explain the inconsistency in
the results of other studies.
Kaine (2004) and Boland et al. (2006) argued that
the adoption of a new agricultural practice depends
on the benets the practice oers. From a farming
systems perspective the benets (and costs) of
adopting a new practice will depend on the extent
to which the new practice integrates with certain
technologies, practices and resources that are already
256 G. Kaine and D. Bewsell
in use. The farm context may include biophysical
characteristics such as climate, soil type and topo-
graphy as well as enterprise characteristics such as
farm layout and shed design. Context may include
factors such as availability and timing of labour, or
technologies such as the type of irrigation or dairy
system in use (Kaine 2004). These technologies,
practices and resources represent the farm context
for the adoption of a new technology and practice
(Crouch 1981). In eect, the farm context denes the
practical benets for the grower of adopting an
innovation and the factors that constitute the farm
context change depending on the innovation under
study. Hence, an understanding of the context for an
innovation provides a basis for drawing inferences
about the reasons why growers do or do not adopt an
innovation.
4. Methods
In the rst stage of the research we used a convergent
interviewing process (Dick 1998) with a sample of
growers from each apple producing region across
Australia to identify the contextual factors inuen-
cing the adoption of IPM. Convergent interviewing is
unstructured in terms of the content of the interview.
The interviewer employs laddering techniques
(Grunert and Grunert 1995) to systematically explore
the rationale behind the decisions and actions of the
interviewee. The power of this interview process lies
in identifying common and complementary patterns
of reasoning among interviewees.
Research and extension sta in the Victorian
Department of Primary Industries provided a list of
fruit growers in Victoria to interview. Care was taken
to include growers with a mix of educational, age and
cultural background operating both large and small
scale enterprises. In addition, a list of contacts from
the appropriate agencies in other Australian states
(New South Wales, Queensland, Tasmania, South
Australia and Western Australia) was sought.
Each contact in the other states was asked to
identify a sample of growers with a mix of educa-
tional, age and cultural background operating both
large and small-scale enterprises. In total 40 apple
growers were interviewed throughout Australia in
the latter part of 2003. A detailed report on the
ndings of these interviews is contained in Kaine
and Bewsell (2003). Some of the data gathered
during the interviews is presented here, however,
names have been changed to preserve the anonymity
of growers.
The results of the interviews were used to design a
mail questionnaire that formed the basis of the
second stage of our research. The questionnaire was
used to collect quantitative data on the factors
inuencing the adoption of a wide range of IPM
and other orchard practices by apple growers. In this
article only the results for IPM practices in relation to
codling moth and pest mites are reported. (For a copy
of the questionnaire, please contact the corresponding
author.)
The questionnaire was piloted with three to six
growers in each state. The questionnaire was mailed,
together with a cover letter and reply paid envelope,
to all apple growers listed by the Australian Apple
and Pear Association in November 2003. A reminder,
in the form of a postcard, was posted 4 weeks later.
The data obtained in the questionnaire were used
to classify growers into segments based on their use of
IPM techniques to manage codling moth and pest
mites. Statistical associations were then identied
between membership of each segment and the
contextual characteristics such as location. We also
tested for statistical associations between membership
of each segment and demographic characteristics such
as age and education. These were Chi-squared tests of
cross-tabulations for categorical data and F-tests for
dierences in means for continuous data (Tabachnik
and Fidell 1989). Segments were identied by apply-
ing Wards procedure with squared Euclidean dis-
tance as the similarity coecient (Aldenderfer and
Blasheld 1984), available in the cluster analysis sub-
programme in SPSS (1988).
5. Results
5.1. Sampling and response rates
The questionnaire was distributed to growers in New
South Wales, Victoria, South Australia, Queensland,
Tasmania and Western Australia. In total, 1313
questionnaires were mailed. Of these, 278 were
completed and returned by growers. Some 243
questionnaires or reminder notes were returned by
respondents who were no longer growing apples or
whose questionnaires were incorrectly addressed.
Based on these data, we estimated that the response
rate to the survey was approximately 28%. Discus-
sions with extension sta from each state indicated
that they believed we had consistently obtained a
response rate of roughly 30% in each state.
5.2. Demographic and enterprise characteristics of
respondents
For the sample as a whole, the average orchard area
was 24 ha, ranging from a minimum of 0.2 ha to a
maximum of 555 ha. There were no statistically
signicant dierences among the states in the average
area of orchards. The average age of the growers in
the sample was 52, ranging from a minimum of 29 to
a maximum of 80. Some 33% of respondents had
primary or some secondary schooling, 26% had
completed secondary school, 16% had completed a
course at a TAFE, technical or agricultural college
and 25% had attended University.
International Journal of Pest Management 257
Approximately 63% of respondents relied on their
orchard for more than 85% of their net income. Only
21% of respondents earned less than 50% of their net
income from their orchards. Some 5% of respondents
had full-time employment o-farm and a further 5%
had regular o-farm employment. Most respondents,
54%, considered themselves debt free or nearly so.
Only 22% indicated that their equity in their orchards
was less than 70%. There were no statistically
signicant dierences among the states in regard to
age or education of growers, their dependence on
their orchards for income, their equity, or the
proportion with o-farm employment.
There were signicant dierences between the
states with regard to the types of fruit crops grown in
conjunction with apples (w
2
61.4 P 5 0.01). In
New South Wales, Victoria and Queensland the
majority of growers grew one or two other tree crops
in addition to apples. In Tasmania and South
Australia most growers only grew one other tree
crop in addition to apples, if any. Most growers in
Western Australia grew three or more tree crops in
addition to apples.
Assessing the extent to which the sample is
representative is a dicult task, as few national
data are available on the demographic characteristics
of apple growers, or on the nancial and physical
characteristics of apple orchards. The age and
education distribution of respondents in our sample
is broadly consistent with those obtained in other
studies such as the Australian Bureau of Statistics
(2003).
5.3. Pest and disease management techniques used
by apple growers
The key factors that were identied by growers during
the interviews as inuencing their potential use of pest
and disease management practices included:
. The climate, which largely dictates the type and
intensity of pest and disease pressures in an
orchard. For example, black spot (apple scab)
was a particularly serious problem in Tasmania
because of the relatively wet conditions experi-
enced in that region. The particular mix of pests
and diseases present, and their intensity, in an
orchard can limit which management options
will be eective in controlling pests and
diseases.
. The topography of the orchard, which can
create microclimates that will have an impact
on the pests and diseases present and the
ecacy of some control options.
. The spatial separation between orchards, which
inuences pest pressures in an area. For
example, many orchards in Queensland and
southern Victoria are geographically distant
from each other. This distance creates a natural
barrier against the movement of pests and
disease between properties. The spatial separa-
tion between orchards appears to result in
considerably lower pest pressures on orchards
that are relatively isolated compared with
others in the same region.
. The mix of tree crops in an area, which can also
contribute to pest and disease pressures. For
example in northern Victoria there was a
concentration of stone and pome fruit. The
presence of stone fruit in the region, which is
particularly susceptible to oriental fruit moth,
adds to the population of this pest in the region.
Similarly, the co-occurrence of stone fruit and
pear crops with apples adds to the codling moth
population in the region.
While climate, topography, orchard isolation and
crop mix determine the type and intensity of pest and
disease pressures growers experience, their manage-
ment of these pressures, i.e. the choice of technology
to deal with these pressures, is determined by
chemical and biological options available for mana-
ging pests and diseases, and the perceived eective-
ness of those options. For many apple pests the only
control options available were broad spectrum
chemicals. For example, selective chemicals or
biological controls have not been developed for
most secondary pests such as weevils. For some apple
diseases chemical eradicants are not available, or are
of limited eectiveness, forcing growers to use cover
spray protectants.
Market requirements were also a major factor
inuencing the options available to growers for
managing pests and diseases. Growers supplying
fresh market fruit often cannot tolerate the same
level of damage, especially cosmetic damage, to fruit
as growers supplying fruit for processing. In addition,
growers supplying fruit for export must conform to
residue limits for those markets. Finally, a key factor
prompting growers to change management options
was the development of resistance to chemicals by
pests.
5.4. Segments for codling moth
Codling moth can be controlled with broad-spectrum
chemicals, with soft chemicals, such as growth
regulators, and by using pheromones to cause
disruption to moth mating. The preferred strategy
in terms of IPM would be to use either a soft chemical
programme or mating disruption to control codling
moth, supported by monitoring and the use of a
publicly available degree day model to predict moth
ights.
Data gathered during interviews indicated that the
development of mating disruption was a signicant
258 G. Kaine and D. Bewsell
step in the control of codling moth in particular, and
for IPM in apple production generally. Mating
disruption allowed growers to reduce their use of,
or even eliminate, broad spectrum chemicals. This
created the opportunity to use IPM methods to
control other pests, particularly pest mites. Codling
moth mating disruption helped growers build up a
population of predator mites to control pest mites
instead of having to rely on acaricides.
Using the ndings from the interviews, questions
were designed to obtain information from apple
growers on the management techniques they used to
control codling moth and pest mites, and their
reasons for using those techniques. The questions
consisted of a series of statements describing each
management technique with a justication for using
that technique (see Table 1). Growers were asked to
nominate which of the statements most closely
reected their situation. Growers could nominate
more than one statement. Using the responses to the
statements listed in Table 1 we classied growers into
four segments with respect to their management of
codling moth. The minimum number of clusters was
identied using the scree test (Aldenderfer and
Blasheld 1984) qualied by inspection for qualitative
dierences between the two clusters aggregated to
form that minimum. The segments are reported in
Table 2.
The rst segment consisted of growers who
appeared to be controlling relatively small codling
moth populations by spraying strategically. They
used the day degree model and monitoring of moth
populations in the orchard to decide when to spray.
The pest pressures in these orchards might be too low
to justify the expense of using mating disruption.
These growers would generally be using
broad-spectrum chemicals though some used a
programme of softer chemicals.
Growers were also asked a series of questions
about their experiences, if any, with using mating
disruption. We found a relatively high proportion of
the growers in segment one had experienced
problems with mating disruption due to edge
eects and diculties with secondary pests (see
Table 3).
Bob is a grower from the rst segment.
Bob has 15 hectares of apples in Uraidla in South
Australia. He sees codling moth as a major pest but
not a major problem. He sprays on the trap count
and usually only sprays about three times in a season.
It is fairly cool so the moth does not get a chance to
develop. Bob tried mating disruption but he found it
was far too costly compared to the cost of using
chemicals like Gusathion
1
[Azinphos Methyl] or
Penncap
1
[Parathion Methyl].
The second segment consisted of growers who used
mating disruption as the basis for control of codling
moth. We believe these growers face moderate to
severe pressure from codling moth. These growers
used chemicals to spray small areas of high infesta-
tion, or hot spots, in the orchard. George, a grower
from the second segment, used mating disruption to
control codling moth.
George has a stone and pome fruit orchard near
Stanthorpe in Queensland. He tried mating disrup-
tion to control codling moth and cut down on the
chemicals he was using as he was experiencing a lot of
problems with pest mites. The predator mites were
being knocked around by the harsh chemicals so to
help him build up the predator mites he trialled
mating disruption. The rst season he used both the
mating disruption and a chemical programme. The
pest pressure came straight down and now he just
uses the mating disruption and is very happy with the
results. He doesnt have a problem with mites now
and doesnt have to struggle with ineective acaricide
sprays.
The third segment comprised of growers who
followed a calendar spraying programme, using
broad spectrum chemicals, because they felt this was
the most cost eective option given the lower pest
pressure they faced. These growers may use monitor-
ing to evaluate the eectiveness of their spray
programme but, generally speaking, would not use
the monitoring to change the timing of sprays unless
they encounter an infestation problem.
The fourth segment consists of growers who
follow a calendar spraying programme using broad
spectrum chemicals because they feel this is the only
eective option given the high pest pressures they
face. These growers may use monitoring to evaluate
the eectiveness of their spray programme but,
generally speaking, will not use the monitoring to
change the timing of sprays.
Table 1. Seven statements on control of codling moth
used to segment respondents.
I calendar spray (e.g., every 10 days) with an
organophosphate (OP) or other chemical because
I have high pest pressure
I calendar spray (e.g., every 10 days) with an
organophosphate (OP) or other chemical because it is
the most cost eective option for me
I use trap catches and/or the degree day model to
determine when to spray with an OP or other
chemical because I have high pest pressure
I use trap catches and/or the degree day model to
determine when to spray with an soft chemical
such as a growth regulator because it is the most
cost eective option for me
Given the pest pressure I have I use mating disruption
in my orchard
I use mating disruption in my orchard but have hot spots
so I spray these areas with an OP or other chemical
sprays
I use mating disruption in my orchard but have hot
spots so I spray these areas with soft chemical sprays
International Journal of Pest Management 259
Nick is an example of a grower from segment four
using a calendar-based spray programme to control
codling moth.
Nick and his brothers run a large orchard near Tatura
in northern Victoria. Nick uses a calendar based
spraying programme to control codling moth. He tries
to avoid using Gusathion
1
[Azinphos Methyl],
instead relying on Folidol
1
[Parathion Methyl].
This spray programme works in well with the calcium
sprays they need to improve the storage life of apples.
The mating disruption costs $450 per hectare and the
Folidol
1
[Parathion Methyl] program only $120 per
hectare. At the moment they dont have a problem
with mites so Nick is very happy with the programme.
Nick has a consultant that comes in every week to
monitor pests and predators. The consultant would
pick up any problems.
There were some signicant dierences among the
segments in the cultivars of apples grown (see
Table 4). A relatively low proportion of growers in
segment three grew cv. Gala apples, a relatively low
proportion of growers in segment four grew cv. Gala
and a relatively high proportion grew cv. Jonathan. A
relatively high proportion of growers in segment two
grew cv. Braeburn.
There was one signicant dierence among the
segments in the types of tree crops grown in
conjunction with apples (see Table 5). A relatively
high proportion of growers in segments three and
four grew pears. Note that a relatively high propor-
tion of growers in segments one and two did not grow
a tree crop other than apples.
In summary, we found statistically signicant
associations between the type of practice used to
control codling moth and contextual factors such as
tree crops grown, orchard layout (edge eects),
intensity of infestations and management of second-
ary pests. We found no statistically signicant
association between segment membership and
Table 3. Proles of codling moth segments on mating disruption (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Codling moth a major problem* 59.6 56.8 35.4 95.8 25.9 50.01
Tried mating disruption* 47.8 85.0 39.6 29.2 40.9 50.01
Mating disruption works well for me*
,{
45.5 67.1 7.7 14.3 23.2 50.01
Problems with mating disruption
due to edge eects*
,{
31.8 8.6 38.5 14.3 11.2 50.01
Problems with mating disruption
due to high pest pressure
{
27.2 10.0 23.1 42.9 6.8 0.08
Problems with mating disruption
due to cost*
,{
13.6 17.1 61.5 42.9 14.8 50.01
Problems with mating disruption
due to secondary pests*
,{
27.3 15.7 38.5 57.1 8.6 0.04
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
{
Indicates proportion of respondents in each segment
who have tried mating disruption.
Table 2. Codling moth segments (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Segment size as a proportion
of respondents who control
for codling moth
23.5 40.5 24.0 12.0 N/A N/A
Calendar spray OP because
of high pest pressure*
0 0 20.8 100.0 143.9 50.01
Calendar spray OP because
most cost eective option*
2.1 4.9 100.0 0 175.5 50.01
Use trap catches/degree day
model to determine when to
spray OP because of
high pest pressure*
100.0 2.5 16.7 12.5 146.5 50.01
Use trap catches/degree day model
to determine when to spray with
soft chemicals because cost eective*
29.8 37.0 0 12.5 25.7 50.01
I use mating disruption* 8.5 40.7 0 0 46.0 50.01
I use mating disruption and spray
hotspots with OP sprays*
14.9 23.5 0 8.3 14.5 50.01
I use mating disruption and spray
hotspots with soft chemical sprays*
14.9 34.6 0 0 31.4 50.01
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
260 G. Kaine and D. Bewsell
orchard size, farm income, operator education or
operator experience (age).
5.5. Segments for mites
Pest mites can be controlled chemically with acar-
icides or biologically using predatory mites. There is
widespread resistance to many acaricides across
Australia and growers are encouraged, if using
acaricides, to limit the number of applications per
season and alternate chemicals from dierent groups.
In Victoria, New South Wales, Queensland, South
Australia and Tasmania pest mites are generally well
controlled by predator mites. Many growers we
interviewed from these states indicated they had not
had a mite problem for several years. For example:
Nick and his brothers run a large orchard near
Tatura in northern Victoria. Generally Nick believes
there is not a lot of mite pressure in the district. We
used to use more acaricides, but since mating
disruption was introduced the number of acaricide
sprays have decreased. Predator mites were intro-
duced 15 years ago [in the late 1980s], but they are
rarely seen in the orchard. Nick believes some
growers in the area still have mite problems because
they use a Gusathion
1
[Azinphos Methyl] based
program that knocks the predator mites around.
Nick will only use Gusathion
1
[Azinphos Methyl]
early in the season for this reason.
Problems can occur with pest mite control when
growers are forced to spray a chemical known to be
detrimental to predator mites to solve another pest
problem. For example:
James runs an orchard in East Shepparton in
northern Victoria growing stone and pome fruit.
James has not had a mite problem for seven years. He
puts this down to the softer chemicals he is using to
control codling moth. He uses a Penncap
1
[Para-
thion Methyl] based calendar programme. Penncap
1
[Parathion Methyl] does not persist in the environ-
ment for very long so has less eect on the predator
mites. This year, because he has introduced bees into
the orchard for owering, he has to go in with
Mavrik
1
[Tau Fluvalinate] to control apple dimpling
bug. This chemical is tough on predator mites so he
will have to watch his trees in order to make sure he
doesnt get a mite problem.
We classied growers into four segments with respect
to their management of pest mites using their
responses to the statements in Table 6 using cluster
analysis as described earlier. The segments are out-
lined in Table 7. The rst segment consisted of
growers who relied on the predator mite population
to control pest mites and who rarely, if ever, used an
araricide. These growers relied on methods to control
pests, such as codling and light brown apple moth
that did not harm predator mites such as mating
disruption, growth regulators and soft chemicals.
From segment one:
William runs a 350 acre orchard near Stanthorpe in
Queensland. He is slowly moving to using mating
Table 4. Proles of codling moth segments on apple cultivars (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Gala* 85.1 91.4 69.8 70.8 12.9 0.01
Fuji 72.3 74.1 58.3 70.8 3.8 0.28
Red Delicious 66.0 74.1 58.3 79.2 4.9 0.18
Pink Lady 76.6 86.4 68.8 70.8 6.5 0.09
Lady Williams 25.5 18.5 27.1 20.8 1.6 0.66
Jonathan* 29.8 44.4 35.4 75.0 14.6 50.01
Braeburn* 12.8 29.6 12.5 4.2 12.1 0.01
Bonza 23.4 22.2 14.6 37.5 4.8 0.18
Sundowner 51.1 51.9 41.7 37.5 2.5 0.48
Golden Delicious 55.3 56.8 52.1 62.5 0.7 0.86
Granny Smith 80.9 76.5 81.3 91.7 2.7 0.43
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
Table 5. Proles of codling moth segments on other tree crops (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Pears* 46.8 35.8 62.5 79.2 17.9 50.01
Nectarines 23.4 21.0 20.8 20.8 0.1 0.99
Peaches 38.3 30.9 39.6 20.8 3.3 0.35
Cherries 19.1 32.1 18.8 41.7 6.8 0.08
Plums 29.8 22.2 33.3 29.2 2.1 0.55
Apricots 21.3 14.8 20.8 12.5 1.7 0.65
Apple only* 29.8 38.3 20.8 8.3 10.0 0.02
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
International Journal of Pest Management 261
disruption on all his apple blocks to control codling
moth. He tried mating disruption in the early 90s
because he had a problem with European red mites.
The sprays he was using to control codling moth
killed the predator mites and the acaricides were not
working very well. Moving to mating disruption has
meant he has not had to spray for red mites.
The second segment consisted of growers who relied
on a predator mite population to control pest mites
but who occasionally used an acaricide to manage
outbreaks on hotspots. The example given above of
Nick and his brothers who run a large orchard near
Tatura in northern Victoria would be representative
of growers from this segment.
The third segment consisted of growers who
routinely used acaricides to control pest mites
because, as a rule, they employed a persistent, broad
spectrum chemical that was harmful to predator
mites. For example:
Matt runs a 280 acre orchard near Manjimup in
Western Australia. Matt nds that the mite problem
on his orchard is variable. He knows he has to be
careful. If he uses Gusathion
1
[Azinphos Methyl] for
weevil control he will have a problem with mites.
However the weevils are a big problem, especially if
they get into the trees.
The fourth segment consisted of growers who used
acaricides because they cannot establish a predator
mite population to control pest mites. Some growers
we interviewed in Western Australia had an occa-
sional problem with pest mites in particular areas in
the orchard, such as alongside a dirt track. For
example:
Frank has a large orchard near Manjimup in Western
Australia. He grows apples, peaches, avocados and
nuts. He has had very few problems with mites, as his
acaricide programme is working well. He alternates
two acaricides to avoid resistance problems. These
are Omite
1
[Propargite] and Vertimec
1
[Abermec-
tin]. Generally the only places he sees mites are along
dusty roads.
Growers in segments one and two were signicantly
less likely to indicate that mites were a major pest
problem than growers in the third and fourth
segments, 13, 26, 68 and 39%, respectively
(w
2
32.4 P 5 0.01). There were no signicant
dierences among the segments in terms of the equity
of growers or the average size of orchards.
There were signicant dierences in the distribu-
tion of segments for mite control across the states (see
Table 8). A relatively high proportion of growers in
Queensland are in segment two while a relatively high
proportion of growers in Western Australia are in
segments three and four. These results are consistent
with the impressions we obtained in our interviews
with growers and with the experiences of extension
sta in those states.
There were signicant dierences in the distribu-
tion of apple varieties grown in each of the segments
for mite control (see Table 9). A relatively low
proportion of growers in segments three and four
grew Fuji, Red Delicious or Braeburn.
There were also signicant dierences across the
segments for mite management in the distribution of
tree crops grown in addition to apples (see Table 10).
A relatively high proportion of growers in segments
three and four grew tree crops in addition to apples,
particularly nectarines and plums.
Finally, there were signicant associations be-
tween membership of mite management segments and
the segments for managing codling moth (see
Table 11). Growers who relied on predator mites to
Table 6. Four statements on managing pest mites used to
classify growers.
The conditions in my orchard mean that I can use
predator mites to keep the pest mites under control
Predator mites in my orchard generally keep the pest
mites under control but occasionally I may have to
use an acaricide* in a particular area or hot spot
I have to use an acaricide* over most or all of my
orchard because I cannot get predator mites established
I have to use an acaricide* over most or all of my
orchard because of the chemicals I am using to
control other pests (e.g. apple dimpling bug or weevils)
*Note that the original statements used miticide rather than
acaricide as the word miticide is more commonly used in Australia.
Table 7. Mite management segments (% of growers in each segment).
Segment one Segment two Segment three Segment four
Segment size as a proportion of
respondents who control for mites
23.9 45.6 17.7 12.8
Use predator mites to control pest mites* 98.1 0 0 0
Use predator mites but occasionally
have to use acaricide** in a hotspot*
31.5 100 10.0 0
Use acaricides** because cannot
establish predator mites*
3.7 0 0 100.0
Use acaricides** because of chemicals used
to control other pests (e.g. weevils)*
0 0 100 31.0
*Denotes statistically signicant dierences in proportions across segments (P 50.05). **Note that the original statements used miticide
rather than acaricide as the word miticide is more commonly used in Australia.
262 G. Kaine and D. Bewsell
control pest mites (segments one and two for mites)
were more likely to have used mating disruption or
soft chemicals to control codling moth (segments one
and two for codling moth). Growers who relied on
acaricides to control pest mites (segments three and
four for mites) were more likely to control codling
moth by calendar spraying a broad spectrum
chemical (segments three and four for codling
moth). This result is consistent with the expectations
of extension specialists.
In summary, we found statistically signicant
associations between the type of practice used to
control pest mites and contextual factors such as
location (climate), tree crops grown and management
of codling moth. We did not nd a statistically
signicant association between segment membership
and orchard size, farm income, operator education or
operator experience (age).
6. Discussion
The adoption of IPM has been said to be driven by
three requirements: the international market, increas-
ing pesticide resistance and environmental responsi-
bility on the part of growers (Urquhart 1999).
However, the conclusion from our review of past
studies into pest and disease management is that there
was little consistency in the ndings of studies into the
adoption of pest and disease management practices.
Some authors have argued that a lack of knowledge
and skills was the key obstacle to the widespread use
of IPM and therefore training and extension was
essential. Others have argued the popularity of
programmes such as farmer eld schools was
primarily attributable to the banning of commonly
used pesticides and fungicides. Concern for the
environment and sustainability had not been identi-
ed as a key factor in the adoption of pest and disease
management techniques in any of the studies we
examined.
The inconsistency in the ndings of previous
studies is understandable if context is considered rst
when investigating growers decision-making about
Table 8. Proles of mite management segments on
location* (% of growers in each segment).
Segment
one
Segment
two
Segment
three
Segment
four
New
South Wales
26.5 49.0 14.3 10.2
Victoria 21.9 54.8 13.7 9.6
Queensland 18.2 63.6 9.1 9.1
South
Australia
33.3 33.3 20.0 13.3
Western
Australia
2.7 29.7 40.5 27.0
Tasmania 50.0 40.9 0 9.1
*Denotes statistically signicant dierences in proportions across
segments (P 50.05).
Table 9. Proles of mite management segments on apple varieties (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Gala 84.9 84.5 75.0 82.8 2.1 0.56
Fuji* 75.9 74.8 50.0 65.5 9.8 0.02
Red Delicious* 79.6 68.0 47.5 51.7 13.1 0.01
Pink Lady 79.6 81.6 85.0 89.7 1.6 0.66
Lady Williams 22.2 18.4 30.0 24.1 2.3 0.51
Jonathan 44.4 33.0 30.0 24.1 4.2 0.24
Braeburn* 24.1 20.4 2.5 10.3 9.7 0.02
Bonza 24.1 24.3 7.5 13.8 6.4 0.10
Sundowner 51.9 48.5 55.0 51.7 0.5 0.91
Golden Delicious 61.1 53.4 45.0 48.3 2.7 0.44
Granny Smith 79.6 74.8 77.5 72.4 0.7 0.87
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
Table 10. Proles of mite management segments on other tree crops (% of growers in each segment).
Segment one Segment two Segment three Segment four w
2
P value
Pears 42.6 43.7 60.0 58.6 5.0 0.17
Nectarines* 14.8 21.4 42.5 41.4 13.9 50.01
Peaches 22.2 33.0 42.5 41.4 5.4 0.15
Cherries 29.6 26.2 20.0 27.6 1.1 0.76
Plums* 20.4 25.2 52.5 48.3 16.8 50.01
Apricots 16.7 15.5 27.5 17.2 2.9 0.40
Apple only* 37.0 31.1 17.5 24.1 1.9 0.60
*Denotes statistically signicant dierences in proportions across segments (P 50.05).
International Journal of Pest Management 263
pest and disease management. We have found that
the particular combination of pest and disease
management techniques a grower employs largely
depends on the mix and intensity of pests and diseases
present in their orchards and the range of control
options available to them. In short, climate, topo-
graphy, orchard isolation and crop mix determine the
type and intensity of pest and disease pressures
growers experience. Their management of these
pressures was determined by the mix of chemical
and biological options available for managing pests
and diseases, and the eectiveness of those options.
These ndings implied that growers followed a
deliberate and systematic process of learning about,
experimenting with, and evaluating management
options within the particular context of their orch-
ards, given the constraints imposed by the realities of
commercial production.
We also found that, once growers had discovered
a combination of management options that was
successful, they would not change that combination
unless forced to do so by circumstances because the
perceived risk of failure was high. Circumstances that
would force growers to change their pest and disease
management included the repeated failure to control
pests due to increasing resistance, the emergence of a
new pest that could not be controlled without
disrupting the control of other pests, the withdrawal
from sale of a key management option or where the
use of an option was declared unacceptable in export
markets.
Our ndings mean that the particular combina-
tion of pest and disease management practices that
best suit an orchard will depend on the particular
circumstances of that orchard. Our ndings also
suggest there is no reason to expect a consistent
relationship between pest and disease management
and orchard size, farm income, operator education
and experience, unless a particular management
technique exhibits scale economies or requires a
formal education qualication to implement. We
did not discover any indication that this was the
case.
Our analysis of the survey of growers was
consistent with those obtained from the grower
interviews. We found weak, if any, relationships
between factors such as the age or education level of
growers and their management of pests and diseases.
The same can be said in relation to orchard size and
grower equity in their orchard. However, we did nd
signicant relationships between contextual factors
such as location and pest intensity and growers
management of pests and diseases that accorded with
the expectations of extension personnel.
These ndings suggest that Jeger (2000) and
Kogan (1998) are correct in attributing the interest
in farmer eld schools in pest management to the
withdrawal of key chemicals from use. The with-
drawal of key chemicals forced producers to search
for alternative management strategies to manage pest
problems. This does not mean that education and
extension does not play an important role in IPM.
Clearly, farmer eld schools are a popular element in
the conduct of that search. Rather, it is important not
to confuse the factors that are motivating change with
the factors motivating participation in a change
programme.
7. Conclusion
We found little evidence to suggest that demographic
factors played an important role in the adoption of
pest and disease management techniques. This result
is consistent with the impressions we obtained in our
interviews with growers. When interviewed, growers
indicated that their orchard circumstances, rather
than their attitudes towards sustainability and the
environment, were the primary determinants of the
particular combinations of pest and disease manage-
ment practices they used. The key factors growers
identied as inuencing their choice of pest and
disease management practices were the climate and
topography of the orchard, the isolation of orchard,
the crop types in the region, the chemical and
biological options available for managing pests and
diseases, and the cost and eectiveness of those
options.
Acknowledgements
We wish to thank David Williams, Jo Vigliaturo and Ben
Rowbottom from the Department of Primary Industries in
Victoria for their help and support with this work. We
would also like to thank all the apple growers who gave up
time to talk to us or complete our questionnaire. This
research was supported by the Australian Research
Council.
References
Abadi Ghadim AK, Pannell DJ. 1999. A conceptual
framework of adoption of an agricultural innovation.
Agr Econ. 21:145154.
Table 11. Association between codling and mite manage-
ment segments* (% of growers in each codling segment).
Mite segments
Segment
one
Segment
two
Segment
three
Segment
four
Codling segments
Segment one 27.3 50.0 15.9 6.8
Segment two 27.2 43.2 19.8 9.9
Segment three 8.7 17.4 34.8 39.1
Segment four 30.0 25.0 40.0 5.0
*Denotes statistically signicant dierences in proportions across
segments (P 50.05).
264 G. Kaine and D. Bewsell
Aldenderfer MS, Blasheld RK. 1984. Cluster Analysis.
Sage University Paper Series on Quantitative Applica-
tions in the Social Sciences, No. 07044. Beverly Hills
and London: Sage Publications.
Australian Bureau of Statistics. 2003. Australian social
trends. Canberra: Author.
Boland A-M, Bewsell D, Kaine G. 2006. Adoption of
sustainable irrigation management practices by stone
and pome fruit growers in the Goulburn/Murray
valleys, Australia. Irrig Sci. 24:137145.
Brewer MJ, Hoard RJ, Landis JN, Elworth LE. 2004. The
case and opportunity for public-supported nancial
incentives to implement integrated pest management.
J Econ Entomol. 97:18821789.
Brough EJ, Frank B, Page F, Lindsay S. 1996. Integrated
mite control in apples in Queensland, Australia. Agric
Ecosyst Environ. 60:129137.
Chaves B, Riley J. 2001. Determination of factors inuen-
cing integrated pest management adoption in coee
berry borer in Colombian farms. Agric Ecosyst
Environ. 87:159177.
Crouch B. 1981. Innovation and farm development: a
multi-dimensional model. In: Crouch B, Chamala S,
editors. Extension education and rural development.
Brisbane: Wiley and Sons.
Dick B. 1998. Convergent interviewing: A technique for
data collection [on line].
Escalada MM, Heong KL. 1993. Communication and
implementation of change in crop protection. Ciba
Foundation Symposium 177. Wiley, Chichester. p. 191
207.
Fernandez-Cornejo J. 1998. Environmental and economic
consequences of technology adoption: IPM in viticul-
ture. Agr Econ. 18:145155.
Fernandez-Cornejo J, Beach ED, Huang W-Y. 1994. The
adoption of IPM techniques by vegetable growers in
Florida, Michigan and Texas. J Agr Appl Econ.
26:158172.
Grieshop JI, Zalom FG, Miyao G. 1988. Adoption and
diusion of integrated pest management innovations in
agriculture. Bull Entomol Soc Am. 34:7278.
Grunert K, Grunert S. 1995. Measuring subjective meaning
structures by the laddering method: Theoretical con-
siderations and methodological problems. Int J Res
Market. 12:209225.
Heong KL, Escalada MM. 1998. Changing rice farmers
pest management practices through participation in a
small-scale experiment. Int J Pest Man. 44:191197.
Jeger MJ. 2000. Bottlenecks in IPM. Crop Prot. 19:787
792.
Kaine G. 2004. Consumer behaviour as a theory of
innovation adoption in agriculture. Hamilton, New
Zealand: AgResearch.
Kaine G, Bewsell D. 2003. Adoption of integrated fruit
production: a qualitative study. AgResearch Ltd.
Kaine G, Bewsell D, Boland A, Linehan C. 2005. Using
market research to understand the adoption of irriga-
tion management strategies in the stone and pome fruit
industry. Austral. J Exp Agr. 45:11811187.
Kogan M. 1998. Integrated pest management: historical
perspectives and contemporary developments. Ann Rev
Entomol. 43:243270.
Malone S, Herbert DA, Jr, Pheasant S. 2004. Determining
adoption of integrated pest management practices by
grain darmers in Virginia. J. Extension. 42. Available
from: http://www.joe.org/joe/2004-august/rb6.shtml.
Norton GA, Adamson D, Aitken LG, Bilston LJ, Foster J,
Frank B, Harper JK. 1999. Facilitating IPM: The role
of participatory workshops. Int J Pest Man. 45:8590.
Nyankanga RO, Wien HC, Olanya OM, Ojiambo PS. 2004.
Farmers, cultural practices and management of potato
late blight in Kenya highlands: Implications for
development of integrated disease management. Int J
Pest Man. 50:135144.
Orr A, Jere P. 1999. Identifying smallholder target groups
in Southern Malawi. Int J Pest Man. 45:179187.
Poubom CFN, Awah ET, Tchuanyo M, Tengoua F. 2005.
Farmers perceptions of cassava pests and indigenous
control methods in Cameroon. Int J Pest Man. 51:157
164.
Rama Rao CA, Srinivasa Rao M, Reddy YVR. 1997.
Predicting the adoption of plant protection measures by
farmers using a logit model. Int J Pest Man. 43:299
302.
Ridgley A-M, Brush SB. 1992. Social factors and selective
technology adoption: The case of integrated pest
management. Human Org. 51:367378.
Rodr guez LC, Niemeyer HM. 2005. Integrated pest
management, semiochemicals and microbial pest-
control agents in Latin American agriculture. Crop
Prot. 24:615623.
Shennan C, Cecchettini CL, Goldman GB, Zalom FG.
2001. Proles of California farmers by degree of IPM
use as indicated by self-descriptions in a phone survey.
Agric Ecosyst Environ. 84:267.
SPSS. 1988. SPSS Users Guide. 3rd ed. Chicago (IL): SPSS
Incorporated.
Tabachnik BG, Fidell LS. 1989. Using multivariate
statistics. 2nd ed. New York: Harper and Row.
Urquhart P. 1999. IPM and the citrus industry in South
Africa. International Institute for Environment and
Development. Report nr Gatekeeper Series No. 86. p 20.
van Mele P, Nguyen TTC, van Huis A. 2001. Farmers
knowledge, perceptions and practices in mango pest
management in the Mekong Delta, Vietnam. Int J Pest
Man. 47:716.
van Mele P, Hai TV, van Huis A. 2002. Inuence of
pesticide information sources on citrus farmers knowl-
edge, perception and practices in pest management,
Mekong Delta, Vietnam. Int J Pest Man. 48:169177.
Waller BE, Hoy CW, Henderson JL, Stinner B, Welty C.
1998. Matching innovations with potential users, a case
study of potato IPM practices. Agric Ecosyst Environ.
70:203215.
Williamson S, Little A, Ali MA, Kiman M, Meir C, Oruko
L. 2003. Aspects of cotton and vegetable farmers pest
management decision-making in India and Kenya. Int J
Pest Man. 49:187198.
International Journal of Pest Management 265