Part 3. Detecting Herbicide Resistance
|
Contents 1. Background 4. Whole plant pot assays
5. Other diagnostic
techniques |
|
1. Background
Reliable tests for
resistance are an essential pre-requisite for the rational implementation of
effective integrated control strategies (See HRAC Guideline to the management
of Herbicide Resistance). Ideally diagnostic tests should be rapid, accurate,
cheap, readily available and give a reliable indication of the likely impact of
resistance on herbicide activity in the field.
Initial suspicion of
resistance usually results from unsatisfactory weed control following herbicide
application. Resistance should not be assumed to be the cause, and other
reasons should be investigated first. Resistance should be considered as a
possible cause when other factors have been eliminated. It is summarized herein
the key principles involved in detecting resistance.
The most important factor determining the ease of detecting resistance
is the degree on insensitivity. When resistance is absolute, and a herbicide
has no visible effect at the recommended rate, detection is easy. With partial
resistance, when some herbicidal effects are seen, detection is more difficult
as resistance is only one of many factors that can reduce herbicide
performance.
2. Field Observation
Accurate field observation is important so that any
reduction in herbicide efficacy can be detected. This may indicate developing
resistance. However, many other factors, apart from resistance, may be
responsible for poor herbicide performance. These include:
a.
Herbicide application factors e.g. inappropriate dose or timing; faulty
spraying.
b.
Soil conditions: e.g. soil moisture; seedbed quality; adsorption.
c.
Climatic conditions: e.g. rainfall patterns; temperature.
d.
Weed factors: e.g. size of weeds; subsequent germination; very high
infestation.
Because so many factors may be responsible for inadequate herbicide
performance, it is often difficult to determine the exact cause of herbicide
failure in the field. Although it is rarely possible to confirm resistance
solely on the basis of field observation and consideration of field observation
and consideration of field records, several factors will point in this
direction. These are:
a. The level of weed control of
other susceptible species. If these
have been controlled effectively, then resistance is a distinct
possibility.
b. The presence of alive plants
adjacent to dead individuals. This
may indicate the presence of resistant individuals, although such situations
can arise through variations in weed growth stage, incorrect application or
through crop shielding.
c. Past experience. If the surviving species has been controlled
successfully by the same treatment in the past, or a gradual decline in control
has been noticed over a period of years, resistance may be responsible.
d. Herbicide history. The repeated annual use of the same herbicide, or
herbicides with the same mode of action, favors selection for resistance (See
HRAC Classification of Herbicides according to Mode of Action).
e. Occurrence of resistance in the
vicinity. If resistance in the same
weed and involving the same herbicide has been positively identified in
adjacent fields or farms, then there is a high probability that resistance is
implicated.
If resistance is suspected, a
sample of seeds (or plants) should be collected from the suspected resistant
weed population for a resistance confirmation test.
3. Seed Collection
The reliability of results based on plant assays is largely
dependent on the quality of the seed sample from which they are grown. Poor
quality seeds will often have low % germination or produce poor plants with
consequent variable response to herbicides.
. collect seeds when the
majority are mature. Collecting
too early or too late is likely to lead to samples with 
low viability. With grass-weeds, e.g.
wild-oats (Avena spp.) such as A. fatua, A. sterilis and
rye-grasses (Lolium spp.) such as L. multiflorum,
the best time is when about 20% of seeds have already been shed.
. collect ripe seeds by
gently rubbing inflorescence over a bag or tray. Seeds of tall weeds, such as wild-oats, are most easily
collected by holding inflorescences inside a large bay and shaking vigorously.
The best technique will vary with species. With grass-weeds it is usually best
to try to collect seeds directly in the field, rather than collect
inflorescences.
. aim to collect over an
area of at least 100m by 50m
within the main problem area, unless the problem is confined to one or more
smaller, very distinct patches. Avoid obvious unsprayed areas. The sample needs
to be representative of the problem field or area, so a few seeds from lots of
heads should be collected. Make a sketch map of area sampled.
. quality is more
important than quantity. Aim
to collect at least a volume of 250 ml of seeds of grass-weeds such as ryegrass
to allow for losses during cleaning. The amount of seed to collect of other
weeds will vary with seed size and ease of collection, but the aim must be to
collect an adequate (several 1000 seeds) sample of ripe seeds.
. do not collect in wet
conditions. Collection is
harder and seeds of some species can become very dormant.
.
beware of rapid heating of freshly collected samples - do not store in
polyethylene bags. Seeds are
best kept in paper envelopes for transport and storage. Staple side and bottom
seams of paper envelopes to prevent them coming unstuck due to moisture from
seeds. Label envelope with name of field, farm and date of collection.
. air dry seeds as soon as
possible after collection.
Small samples can be dried in the envelopes by simply standing them on end
with the flap open, and shaking the envelope daily. Larger samples are best
dried in trays placed in a dry, well ventilated, but not windy, environment.
Seeds of most species should be dry within about a week.
. clean samples to remove
poor quality seeds. The best
technique for cleaning samples will vary with species but sieving to remove
large pieces of plant debris and air flow to remove lighter seeds are
appropriate for many species.
4. Whole Plant Pot Assays
The most widely
used test for resistance involves growing plants from seeds collected from the
suspect field, and spraying them with herbicides applied either at a single
discriminating dose, or a range of doses. Such assays are usually conducted in
a glasshouse or controlled environment chamber. Assessments usually involve
visual assessments of mortality or plant vigour, or measurements of fresh or
dry weight of foliage.
. An essential component
of all resistance assays is the inclusion of an appropriate susceptible
reference population. Susceptible
standards should be chosen with care, to ensure that they are truly
representative, and not atypically sensitive or insensitive to the herbicide
under evaluation. Inclusion of several susceptible standards is recommended,
especially when resistance is partial, as this will provide information on the
background range of responses to herbicides.
. Statistical advice
should be sought to ensure that the experiment design and replication is
appropriate. Experiments that include
populations with varying levels of resistance, often introduce a large amount
of variability into the resulting data.
DOSE
RESPONSE EXPERIMENTS
. In initial studies it is
preferable to use a range of doses to obtain a response curve. This enables the degree of resistance to be better quantified
by calculating the ratio of doses required to produce the same effect in
resistant and susceptible populations. Usually the dose required to give a 50
(70 reduction in the measured parameter (usually foliage weight or number of
surviving plants), relative to the untreated control is determined (Figure 1).
. Ratios of these
estimates, (variously termed ED50, GR50, LD50 or 150), relative to that of a
susceptible population, provide a resistance index (RI) which enables the
degree of resistance to be described relatively simply.
. To obtain a good
estimate of ED50 the dose range should be relatively wide and at least six
doses are needed. It is usually best that
each dose is twice the preceding dose in the range (e.g. 10, 20, 40, 80, 160,
320 g a. i./ha). The dose range used should include doses both below and above
the field recommended rate as herbicides are normally more active under
greenhouse conditions.
|
|
|
SINGLE DOSE RESISTANCE ASSAYS
. Once dose response information
has been obtained, it is often possible to use a single (or two or three)
discriminating dose(s) in future screening assays, which allows many more
populations to be tested as fewer pots per population are needed. With some forms of resistance, such as most cases of
resistance to triazine herbicides, resistance tends to be absolute. In such
cases, resistance is easy to identify and choice of dose is not critical - so
long as it kills susceptible plants. When resistance is partial, more care is required
in choosing the most appropriate single dose.
A 'ring test' involving 16
organizations in 8 European countries has recently been undertaken to evaluate
the consistency of resistance screening tests in order to improve the
standardization of testing procedures (Moss et al., 1998). As a
consequence of this study, the following recommendations were made:
RECOMMENDATIONS
. Ensure adequate seed
supplies are available and clean them to remove poor quality seeds. Poor quality, insufficient, seed samples are likely
to result in poor quality plants which may be more, or less, susceptible to
herbicides.
. Prior to spraying
achieve well matched plants in terms of growth stage and vigour by sowing pre- germinated seeds or by sowing plenty
of seeds and thinning down to a constant number per pot.
. Do not rely solely on
sub-irrigation for watering if soil-acting herbicides are being used as this will prevent herbicides being moved down into
the plant rooting zone.
. If a single dose assay
is used, the best single herbicide dose is likely to vary between individual
testing centers and can only be
determined by preliminary experimentation. Herbicide activity will be affected
by numerous factors, but the most important factors are likely to be the soil
organic matter level (for soil acting herbicides) and the growing conditions
(especially light and temperature).
. Use susceptible and
resistant standard reference populations in every assay. Ideally, different testing centers should use
identical standards for each species. Do not assume that all susceptible
populations are equally susceptible to all herbicides. Choose standards
carefully and consider availability of seeds in the longer term.
. In single dose assays,
aim to achieve an 85-95% reduction in foliage fresh weight for the
susceptible standard. Too high or
low a level will reduce the sensitivity of the assay.
. Aim for <50%
reduction in foliage fresh weight for any resistant standard. If appropriate, include both a highly resistant
(expected 0% reduction) and partially resistant (about 50% reduction)
standards. Inclusion of only a highly resistant standard will not allow the
relative herbicide efficacy between subsequent assays tests to be
determined.
. Ideally record foliage
fresh weight as an objective assessment of herbicide activity, when full
effects of the herbicide are evident on the susceptible standard. The time from spraying to assessment will vary with
herbicide used, weed species and environmental conditions. With many weeds and
herbicides, a three-week time span between spraying and assessment is
appropriate for plants kept in glasshouse conditions.
. Visual assessments may
be a suitable alternative and are
certainly much quicker than weight assessments. If visual assessments alone are
conducted, record foliage weights for the susceptible and resistant standard
reference populations. This data can be used to check on the accuracy of the
visual assessments and the consistency of results between subsequent
assays.
. Regardless of how the
screening assay is conducted, the basis on which resistance is assigned should
be stated. This is particularly important
where populations show marginal or partial resistance.
. Comparison of results
obtained from different testing centers should be done with care, especially when resistance is partial, rather than
absolute. Consistency between assays conducted at any one center is likely to
be better than between centers.
4. Other Diagnostic Techniques
Other
diagnostic techniques have been developed for detecting specific forms of
resistance. These include pots tests using field collected plants, Petri-dish
germination assays, chlorophyll fluorescence, leaf disc flotation and enzyme
sensitivity assays. These have been reviewed by Moss (1995). Most of the
principles outlined above are also relevant to these other techniques. However,
the glasshouse pot assay is likely to remain the most appropriate single test
for resistance as herbicide application and activity mimic what happens in the
field. In addition pot assays can detect resistance regardless of mechanism - a
very important attribute.
More specific assays may be
quicker and more precisely identify the mechanisms responsible, but their very
precision may be a limitation, especially where multiple mechanisms of
resistance exist. In addition, care must be taken in interpreting results from
methods which involve using herbicides in ways totally different to field
applications.
6. Interpretation
of Results
It is
important to recognize the fact that plants or seeds collected for resistance
tests usually represent a biased sample.
How representative they are of the entire field depends on the method of
sampling and the proportion of plants that survived treatment in the field. If
seed samples were collected from a few surviving resistant plants, when the
majority of susceptible plants were killed, then any test result will overstate
the degree of resistance currently present in the entire field population. This should not be viewed as a limitation of diagnostic
assays, but a positive attribute, as it enables resistance to be detected at an
early stage of development, when it is easier to take action to prevent
the situation getting worse.
. With results from dose response experiments, the higher the resistance index (ratios of ED50 values relative to that of a susceptible population), the greater the level of resistance (Table 1). Small resistance indices (e.g. 2-3) can occur between normal susceptible populations, so these should be interpreted with care, regardless of statistical significance. With highly resistant populations it may not be possible to obtain an ED50 value and so a precise resistance index cannot be calculated.
|
Table 1. Results of a
glasshouse dose response investigating the effect of fenoxaprop on four populations
of Alopecurus
myosuroides.
|
When resistance is absolute,
interpretation is relatively easy as plants are either likely to be alive
(resistant) or dead (susceptible) over a wide dose range. In such situations simply
expressing the proportion of plants surviving treatment is likely to be
appropriate, although how representative the tested sample is of the entire
field population must be born in mind. When resistance is partial,
interpretation is more difficult (Table 2). Statistical comparisons, while
essential for research studies, are not necessarily appropriate in routine
screening tests.
|
Table 2. Results of a
glasshouse pot-screening assay in which a single dose of fenoxaprop (55 g a.i./ha) was
applied to four Avena
fatua populations. |
|
|
Population |
% reduction in foliage weight* |
|
W(susceptible) X Y Z |
93% 7% 68% 84% |
* = relative to untreated control pots for same
population.
Interpretation: The
susceptible standard, population W, was well controlled by this dose of
herbicide. Control of population X was very poor indicating that it was
resistant. Population Y was partially controlled, indicating partial
resistance. There appeared to be a marginal difference between the susceptible
standard (W) and population Z. Further studies would be needed to determine
whether this difference had any relevance in the field.
. With single dose assays, one
classification system that can be used to assign different degrees of
resistance is a * rating system which encompasses the concept of varying
degrees of resistance at the population level. The original system required the
inclusion of three reference populations, but the revised system (Clarke, Blair
& Moss, 1994) requires the inclusion of only two reference populations, one
susceptible and one resistant, which are included in every test.
. Results from resistance screening
experiments should be related to the herbicide performance in the sampled
fields. It then becomes possible to use diagnostic test results to predict, at
least to some degree, the likely impact of resistance on herbicide performance
elsewhere.
It is generally
concluded that one of the primary aims of integrated weed control must be to
try to prevent herbicide- resistance developing. However, if this is
unsuccessful, it is vital that resistance to herbicides is detected as early as
possible so that resistance management strategies can be implemented. If
resistance becomes an acute, whole farm problem, then control options are more
limited and greater expense and effort will be almost inevitable. Confirmation
of resistance can result in substantial changes to the farming system e.g.
changes to crop rotation, cultivation practices and the use of more expensive
herbicides. Therefore it is essential that resistance tests are conducted
properly if reliable and meaningful results are to be obtained. It is hoped that
these guidelines will help achieve this goal.
Source
Dr Stephen Moss
IACR-Rothamsted
Harpenden
Herts AL5 2JQ
UK
http://www.plantprotection.org/HRAC/detecting.html
Further information
Dr.
David Vitolo
Syngenta R-1004, Basel, CH-4002, Suiza
http://www.plantprotection.org/HRAC/