Insect management in fababeans and canola recent research

Take home message

Compensatory capacity of canola supports the use of less conservative aphid thresholds, and increased consideration of natural enemies in controlling outbreaks.

Beatsheet and sweep net sampling of faba beans both severely underestimate the density of larvae, and particularly the smaller larvae. A visual inspection of faba bean terminals, flowers and buds is critical to improved estimates of smaller larvae.

Whilst important, insect damage does not constitute the majority of defective grain.

Outcomes and recommendations. Management of aphids in canola

  1. Simulated damage through the removal of raceme terminals by cutting provides an adequate way of simulating aphid damage.
  2. Trials conducted on the Darling Downs in 2013, and on the Liverpool Plains in 2014 show identical trends in terms of crop compensation for simulated aphid damage. Consequently, we can have some confidence that the conclusions drawn from these trials will have application to canola crops across a wide range of northern region growing conditions.
  3. Removal of the flowering portion of podding racemes (23 days post first flower) did not impact on yield, probably because the flowers removed would not have set harvestable pods.
  4. Simulated damage to the raceme during the first weeks of flowering had only minimal impact on final yield except where extreme damage was enacted (66% of or the entire raceme removed on every plant).
  5. Compensatory capacity and minimal maturity delay following damage to racemes suggests greater opportunity to harness the benefit of natural enemies in controlling aphid outbreaks.

Aphid populations are extremely difficult to work with in the field. Manipulating densities, frequency and persistence of infestations are major constraints to getting trial outcomes. Therefore, simulated aphid damage is the only viable way to apply consistent treatments across a replicated trial.

The purpose of the simulated aphid trials reported here was to assess the impact of differing levels of damage at different stages of crop development. The trials were designed to evaluate how much compensatory capacity canola has to recover yield if damaged by aphids at different stages of crop development. Understanding when canola is most susceptible to aphid-type damage is the first step in determining the need for, and timing of, aphid control in canola.

Darling Downs trials were conducted in dryland canola (43Y85).

Figure 1. The two methods of inflicting simulated aphid damage. Bagging racemes to prevent normal development, used in 2013 only (left), and cutting racemes (right). 

Figure 1. The two methods of inflicting simulated aphid damage. Bagging racemes to prevent normal development, used in 2013 only (left), and cutting racemes (right). 

Experiment 1.  Bagging canola racemes

In the first experiment canola raceme development was inhibited by the placement of bags over the developing main stem raceme to limit development.  Three treatments and an untreated control were applied to randomly assigned plots. The primary raceme on all plants was bagged at first flower (treatment 1), 7 day later (treatment 2) and then 14 days after first flower (treatment 3).

Analysis of grain yields showed that all of the bagged treatments yielded significantly less than the control but were not different from each other regardless of treatment timing (Table 1). The maturity of the plots at the end of the season was very similar to the control being possibly 3-4 days behind in terms of the rate of plant senescence. This results shows that the later flowers (those prevented from developing at +7 and +14 days) contributed little to the final yield, or that soil moisture limited continued growth/compensation in the treated plots.

Table 1. Plot yields from experiment 1 where canola raceme development was limited by bagging at 7 day intervals from first flower.

Treatment

Yield

SE

Control (unbagged)

1293  a

63

Treatment 1- Raceme covered first flower

988  b

18

Treatment 2 - Raceme covered first flower +7 days

989  b

62

Treatment 3 - Raceme covered first flower +14 days

949  b

25

Means with different letters are significantly different (P<0.002). LSD 0.05% = 146g

Experiment 2.  Increasing severity of cutting flowering racemes

A second experiment was laid out in the same canola field to test the effects of damage at different intensities during early flowering. Four treatments were applied with an undamaged control.

The damage applied to the plots at 14 days post first flower resulted in a significant yield loss only for the most damaging treatment;  where all developing raceme terminals were cut (top 7 nodes) (Table 2). 

The rate of crop senescence was observed to be marginally delayed in the two most severe cutting treatments. Crop maturity of treated plots was delayed by approximately 3-4 days compared with the controls.

Table 2. Plot yields from experiment 2 where canola racemes were cut with increasing severity at 14d post first flower.

Treatment

Yield

SE

Control (undamaged)

1137  a

108

Terminal of raceme cut

1173  a

98

Top 3 nodes of raceme cut

1133  a

65

Top 5 nodes of raceme cut

1102  a

89

Top 7 nodes of raceme cut (nearly all)

841  b

111

Means with different letters are significantly different (P<0.005). LSD 0.05% = 242g

Experiment 3.  Damage to flowering racemes at late flowering-pod filling

The third experiment removed just the flowering portion of filling racemes at 23 days after first flower. In the treated plots, the terminals of  0, 10, 50 or 90% of the racemes in the plot were damaged. These treatments were designed to simulate the late terminal infestations commonly observed in late spring.

This damage to the flowering terminals during late flowering –pod set, had no significant impact on grain yield (Table 3). There was no delay in crop maturity and no significant difference in the rate of crop senescence observed between treatments.

Table 3. Plot yields from experiment 3 where increasing proportions of racemes were damaged during late flowering – pod set.

Treatment

Yield

SE

Control

1246  a

70

10% of racemes terminal removed

1159  a

37

50% of racemes terminal removed

1193  a

65

90% of racemes terminal removed

1204  a

89

Treatments were not significantly different (P<0.05). LSD 0.05% = 161g

Experiment 4. Liverpool Plains 2014

To further test the impacts of simulated damage on flowering canola a second series of experiments were conducted at 8 sites near Spring Ridge on the Liverpool Plains. For this experiment three treatments were applied with an undamaged control at each site.

The damage was inflicted during the early stages of flowering (within 7-10 days of first flower) at five sites during early August and a further 3 sites that began flowering during late August. The results from this experiment are still incomplete as an analysis of oil quality and seed characteristics remains to be completed but harvested grain weights indicated again that main treatment to suffer significant damage was the plots where the entire raceme was removed leaving plants to regrow completely new flowering structures.

The rate of crop senescence was observed to be marginally delayed in the more severe cutting treatments being 3-5 days behind the undamaged controls even for the most severe cutting treatment.

 Figure 2. Images of treatments showing at treatment and compensation by the crops 15 days later

Treatment where 2/3 of raceme flowers removed (top) and 15 days later showing compensation (bottom).

Treatment where entire raceme is removed (top) and 15 days later showing compensation (bottom).

Figure 2. Images of treatments showing at treatment and compensation by the crops 15 days later

Figure 3. Mean canola yield for treatments enacted soon after first flower during early August (5 sites) and late August (3 sites). Bars denote treatment SE. 

Figure 3. Mean canola yield for treatments enacted soon after first flower during early August (5 sites) and late August (3 sites). Bars denote treatment SE.

 Figure 4. Aggregate treatment yields across ALL sites near Spring Ridge NSW, expressed as a percentage of the control.

Figure 4. Aggregate treatment yields across ALL sites near Spring Ridge NSW, expressed as a percentage of the control.

Conclusions and observations

The bagging of canola racemes during early flowering had a significant effect on yield potential (experiment 1). In contrast, cutting the racemes (experiment 2) only had an impact on yield for the most severely damaged treatment that had all raceme terminals from the top 7 nodes disrupted (essentially all flowering racemes on the plants). The yield of the other treatments was not significantly affected, although the second most severe damage treatment did trend towards a lower yield.

The difference in crop response to bagging and cutting is probably caused by the plants continuing to develop racemes under the bag for a significant part of the flowering period rather than deploying assimilates to newer compensatory racemes from lower down in the plant canopy. This contrasted with the plants where damage was inflicted by cutting, where compensatory growth of remaining racemes or the initiation of new racemes from adjacent nodes occurred rapidly after damage was inflicted.

Late damage to flowering-podding racemes did not have any impact on crop yield. The flowers removed from the racemes at this stage of flowering would not have set viable pods, so their loss did not impact on yield.

Whilst the physical damage inflicted during these experiments is different to that which could be expected during an aphid infestation the results suggest that canola has an excellent capacity to compensate for crop damage and that current spray thresholds of around 10% raceme infestation maybe more stringent than necessary.

Given the capacity for compensation and the regular presence of effective aphid natural enemies in canola we contend that there is an excellent opportunity rely on and allow time for biological control by aphid parasitoids and ladybirds to take place during the first weeks of flowering. A delay in enacting a spray decision at the 10% infestation level could be considered to be a low risk and allow time for biological control.  If natural enemies were ineffective, spraying on an increasing level of infestation to the 20-25% level would be unlikely to result in irrecoverable crop damage. Similarly late infestations of aphids are also unlikely to pose a damage threat to canola as the associated raceme disruption mainly affects flowers that contribute little to final yield.

During the 2015 we hope to conduct a study to determine a better linkage between the simulated damage in these experiments with that caused by aphids in highly controlled small plot experiments so that we will be able to make a definitive recommendation on aphid threshholds. In the interim our work suggests that the use of the 10% raceme infestation as a spray threshold is very conservative.

Insect pest management in faba beans

In 2014, the warm autumn provided an opportunity to look at a number of insect pests in faba beans. The results reported here are preliminary, and require further investigation before definitive recommendations are determined. However, for some of these pests, this information is more than has been available to date to guide management decisions.

Helicoverpa

Helicoverpa is the most significant pest of faba beans in the northern region, occurring in most seasons from flowering to maturity. There is much debate over whether the threshold of 2 larvae/m2 (by beatsheet sampling) is providing adequate control. There is no evidence that the helicoverpa threshold is based on trial work, rather it is likely to be a ‘best guess’. It is also likely that the threshold used in the north was originally 2 larvae per 10 sweeps (adapted from the south). The only threshold for helicoverpa in faba beans that may have been derived from trial work is the Western Australian threshold, which is: 90 kg yield loss per larva per 10 sweeps.  Both these thresholds appear to focus on yield loss, and do not consider quality. It is highly likely that in faba beans the quality threshold (defective grain) is reached before significant yield loss (kg/ha) occurs. The threshold clearly requires research attention.

A simple conversion from sweep net sampling to beatsheet sampling is not currently possible because the two methods are not equally effective at collecting different larval sizes (Figure 5). Both methods are providing very low estimates of what is actually present in the crop, and are particularly poor at estimating the number of small larvae. This result is problematic, and may be a significant contributor to the perception of the threshold as being unreliable. If you cannot get a reliable estimate of how many larvae (and what size larvae) are in the crop, then it is impossible to effectively manage. The overall efficiency of the sampling (total larvae) will be heavily influenced by whether the population is biased towards smaller or larger larvae. If it is largely smaller larvae (as this crop was with 80% of larvae S-SM), then the sampling will severely underestimates the population. Being able to detect smaller larvae, before they start to damage pods is critical to preventing both quality and yield loss in faba beans.

Figure 5. Helicoverpa larvae in flowering faba beans are sampled with differing efficiency. Larger larvae are more easily dislodged with both methods, but smaller larvae are more difficult to assess accurately. The beatsheet sampling method is providing a better estimate of larval density, but is still poor overall.

Figure 5. Helicoverpa larvae in flowering faba beans are sampled with differing efficiency. Larger larvae are more easily dislodged with both methods, but smaller larvae are more difficult to assess accurately. The beatsheet sampling method is providing a better estimate of larval density, but is still poor overall.

The smaller larvae are more difficult to dislodge with both the sweep net and beatsheet because they are ‘entrenched’ in the reproductive structures and terminal (Figure 6).

Figure 6. Location of eggs and larvae on different structures in flowering faba beans. 

Figure 6. Location of eggs and larvae on different structures in flowering faba beans.

To ensure that these smaller larvae are not ‘ignored’ in sample estimates, it is important to visually examine these reproductive structures and the terminal. A visual inspection of the upper portion of the plant (flowers and terminal) should be sufficient (Figure 7).

Figure 7. Smaller larvae are more abundant in the top part of a flowering faba bean crop (flowers, buds and terminal) than the lower portion (pods). 

Figure 7. Smaller larvae are more abundant in the top part of a flowering faba bean crop (flowers, buds and terminal) than the lower portion (pods).

Defective grain and the contribution of helicoverpa

One of the concerns that has been raised repeatedly in terms of helicoverpa management in faba beans is the issue of defective grain. Because it is likely that the nature of pod feeding (repeated minor penetration of pods) results in more damaged grain than significant yield loss; insect damage is suspected of being a major contributor to defective grain. To determine if this was the case we analysed the data from 120 receival notes from faba bean deliveries in NSW between 2010 and 2012. We wanted to know if insect damage made up a major proportion of the overall defective grain.

The results of this analysis were:

  1. the main contributor to defective grain was broken/damaged grain (2% of all seed)
  2. insect damage was the second largest contributor (1.3% of all seed)
  3. other categories in rank order were weathered (1.2%), shriveled (1.1%) and poor colour (0.8%).

These categories were those used at the receival depots. Further analysis to consider what field, harvest and storage practices may be influencing the level of defective grain (non-insect) is warranted as overall these contribute significantly more to defective grain than insect damage.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the author would like to thank them for their continued support.

We thank Sam Simons and Jim Hunt who assisted us with maintaining trials sites in the Spring Ridge region.

Wade Bidstrup generously provided a portion of his faba bean crop for much of this work, for which we are extremely grateful. A number of agronomists on the Darling Downs and Northern NSW provided seed samples and information on pest infestations and trial sites; we thank them for their interest and valuable support. Kerry Bell (DAFF Qld) provides biometry support for our trial work and analysed the receival data.

Contact details

Melina Miles
Ph: 0407113306
Fx: 07 46881199
Email: melina.miles@daff.qld.gov.au

Paul Grundy
Ph: 0427929172
Email: paul.grundy@daff.qld.gov.au

GRDC Project Code: DAQ00153,