Prof. Jerry Cross. Paul Bennett, Colin Corfield

Transcription

1 Project title: Optimum treatment timing to reduce overwintering codling moth populations 9- Project number: TF 189 Project leader: Prof. Jerry Cross Report: Year 2 report /11 Previous report: Annual report 9/ Key staff: Prof. Jerry Cross Dr Michelle Fountain G Arnold BA, MSc, CStat (Biometrician) Location of project: East Malling Research, New Road, East Malling, Kent, ME19 6BJ. Industry Representative: Paul Bennett, Colin Corfield Date project commenced: 1 April 9 Date project completed (or expected completion date): 31 March 11 Agriculture and Horticulture Development Board 11. All rights reserved.

2 DISCLAIMER: AHDB, operating through its HDC division seeks to ensure that the information contained within this document is accurate at the time of printing. No warranty is given in respect thereof and, to the maximum extent permitted by law the Agriculture and Horticulture Development Board accepts no liability for loss, damage or injury howsoever caused (including that caused by negligence) or suffered directly or indirectly in relation to information and opinions contained in or omitted from this document. Copyright, Agriculture and Horticulture Development Board 11. All rights reserved. No part of this publication may be reproduced in any material form (including by photocopy or storage in any medium by electronic means) or any copy or adaptation stored, published or distributed (by physical, electronic or other means) without the prior permission in writing of the Agriculture and Horticulture Development Board, other than by reproduction in an unmodified form for the sole purpose of use as an information resource when the Agriculture and Horticulture Development Board or HDC is clearly acknowledged as the source, or in accordance with the provisions of the Copyright, Designs and Patents Act All rights reserved. AHDB (logo) is a registered trademark of the Agriculture and Horticulture Development Board. HDC is a registered trademark of the Agriculture and Horticulture Development Board, for use by its HDC division. All other trademarks, logos and brand names contained in this publication are the trademarks of their respective holders. No rights are granted without the prior written permission of the relevant owners. Agriculture and Horticulture Development Board 11. All rights reserved.

3 AUTHENTICATION We declare that this work was done under our supervision according to the procedures described herein and that the report represents a true and accurate record of the results obtained. [Researcher A Name] [Position] [Organisation] Signature... Date... [Researcher B Name] [Position] [Organisation] Signature... Date... Report authorised by: [Organistion A's authority Name] [Position] [Organisation] Signature... Date... Agriculture and Horticulture Development Board 11. All rights reserved.

4 CONTENTS Grower summary 1 Headline 1 Background 1 Financial benefits Action points Science section 7 Introduction 7 Materials and methods 7 Results Discussion 22 Conclusions 27 Acknowledgements 43 References 43 Agriculture and Horticulture Development Board 11. All rights reserved.

5 GROWER SUMMARY Headline Use pheromone traps, egg monitoring and a temperature based forecasting model for timing codling moth sprays and focus on good control of the 1 st generation in e and y to minimise population carry over to the next year. Background and expected deliverables Codling moth is still the most important apple pest in the UK. Despite regular use of insecticides through the season to control it and prevent the damage the larvae cause, populations persist from year to year to give rise to new attacks. Work is required to study the egg-laying dynamics of codling moth through the season in relation to pheromone trap catches and forecasts by the RIMpro-Cydia model, to find when treatments are best applied to reduce overwintering populations. The results should guide control strategies for long-term reduction of moth numbers. Summary of the project and main conclusions The seasonal dynamics of codling moth were studied in two unsprayed areas (each of ~ trees) in infested apple orchards near Faversham and Lamberhurst, Kent. Two standard sex pheromone baited delta traps were used to monitor males and two pear ester kairomone (DA2313) bated delta traps to monitor males and females (early to mid-ober). Each week, a count was made of the codling moth eggs and larvae on 3- fruit clusters per orchard. The approximate date of egg hatch for each larva collected was back-calculated from temperature records. Larvae migrating to pupation sites were monitored from mid- to tember at fortnightly intervals using corrugated cardboard bands around the tree trunks. Pupation success was quantified to determine contributions to the infestation the following year. The usefulness and accuracy of the RIMpro-Cydia computer based codling moth forecasting model were investigated. Seasonal dynamics of codling moth and targeting of sprays There was a strong 1 st codling moth generation between mid- and end of y and a weak incomplete 2 nd generation between early ust and tember or early ober at both sites in both years. Agriculture and Horticulture Development Board 11. All rights reserved. 1

6 Temperature sum calculations indicate that only those individuals completing their 1 st generation development and laying eggs before the end of y have any chance of completing a 2 nd generation in the same year. In 9 and the proportion of the population that did this was zero. 1 st generation eggs laid after the end of y did not complete their development. Thus, the 2 nd generation, though damaging to the crop, does not complete its development or contribute to the following year s population. Sex pheromone traps This study confirms that sex pheromone traps should not be relied on solely for monitoring codling moth populations and timing sprays. They gave a broad indication of the periods of codling moth activity, but could not be used to predict the occurrence of eggs in any qualitative or quantitative way. Early catches of 1 st generation males preceded the first appearance of eggs by 4- weeks in 9 but by less than two weeks in. Importantly, the ratio between the numbers of male moths caught in the sex pheromone traps and the numbers of eggs found on fruits in the two orchards over the two years varied greatly and was highly inconsistent. For the 1 st generation, the ratios were 93:37, 88:13, 1:237 and 138:19 for the two orchards in the two years, respectively. Furthermore, the pheromone traps scarcely indicated the 2 nd generation flights and attacks. For the 2nd generation, the ratios of males to eggs were 4:11, 9:, 7:122 and 14:6 for the two orchards over the two years, respectively. At best only a few moths were captured in ust and tember, even at one site (Teynham) in where a heavy 2 nd generation attack occurred. Based on the UK data provided by this study, if pheromone traps were to be relied on, a much lower threshold, say of two moths per trap per week, would need to be adopted for the 2 nd generation which can be very damaging as the fruit is ripening and becoming more susceptible to attack and larval penetration. Sex pheromone traps are relatively quick and easy to use and do give valuable information on the general occurrence of the pest, including flight periods and activity, but catches need to be interpreted with care and in conjunction with other monitoring Agriculture and Horticulture Development Board 11. All rights reserved. 2

7 and forecasting methods. If relied on alone with the existing low threshold of moths per trap per week and perhaps a lower threshold of two moths per trap per week for the 2 nd generation, excessive and not very well targeted application of sprays is likely to result, which will do an adequate job but often at higher cost than necessary. In addition, the reduction in the carryover of the population to the following year will not be optimised. Pear ester traps o The pear ester kairomone traps performed rather similarly to the sex pheromone traps in terms of capture of 1 st generation males, though numbers of males captured in the pear ester traps were much lower than captures in the sex pheromone traps; also the peaks and troughs of capture with the two trapping methods did not coincide. The start of 1 st generation pear ester trap catches were a week or two later than the sex pheromone trap catches, probably due to reduced attractiveness. The pear ester traps did not give any indication of the 2 nd generation (of males or females) and this was a serious failing. o The main advantage of the pear ester traps is that they capture females, though numbers captured in this study were small and erratic, and no 2 nd generation females were caught at all. The catches of females did roughly coincide with the periods of egg-laying, but the erratic nature and small size of the catches could not be relied on to give a clear indication of the times of egg laying or of the numbers of eggs. Again there was no consistent ratio between the numbers of females captured and the numbers of eggs. The moths are easy to sex by the presence of a characteristic dark patch on the undersides of the forewing of males which is absent in the females. o The failure to catch 2 nd generation females was probably due to a lack of attractiveness of the pear ester lures and increasing competition with the volatiles that were produced by the fruits as they ripened. The use of the pear ester traps is worthy of further investigation but this study did not indicate their use would make a substantial contribution to improving codling moth monitoring. Egg monitoring Regular, direct counting of eggs on fruitlets in situ is currently probably the only fully reliable and accurate way of determining the timing and need for sprays against codling moth. Agriculture and Horticulture Development Board 11. All rights reserved. 3

8 Direct egg monitoring is time consuming and requires diligence and good eye sight, but is not difficult. Observations ideally would need to be done weekly with a sample size of ideally at least 6 fruitlets per orchard per sampling occasion. Targeting the sampling to where codling moth damage is most likely to occur on fruits exposed to sunshine at the top of the tree will markedly improve efficiency. The sample size could be reduced when egg numbers are high. The time required to examine 6 fruitlets is at least an hour, so such sampling is costly. However, codling moth control is probably costing growers at least 3 per ha per season and better targeting and timing of sprays might reduce this by at least per ha per season, so investment in egg monitoring is justified. Forecasting models In view of the limited value of the information provided by sex pheromone trap monitoring, computer based phenological simulation models are an important source of information for interpreting the seasonal dynamics of codling moth. The RIMpro-Cydia model is the most sophisticated model available. It did give broadly correct predictions of the 1 st generation seasonal dynamics, taking into account well established day-degree air temperature sums > C as well as dusk temperatures > C which are known to be important for egg laying activity. However, the RIMpro-Cydia model failed to predict the occurrence of the 2 nd generation attack in both years, a particularly serious failing at Teynham in, where large numbers of eggs were laid throughout ust and into tember which resulted in a significant larval attack. Temperature based models are sensitive to air temperatures and the accuracy of grower operated weather stations needs to be scrutinised Control tactics The seasonal dynamics of codling moth in the UK has implications for control tactics. Control measures to maximise the reduction of the carry over of the populations to Agriculture and Horticulture Development Board 11. All rights reserved. 4

9 the following season and avoid the development of resistance should focus on control of the 1 st generation. A high standard of control thus needs to be maintained throughout e and y. During this period it is important to use a range of insecticides with different modes of action to minimise the risk of resistance development. Use of codling moth granulovirus as part of the strategy against the 1 st generation is important. Some UK growers may be under the misapprehension that codling moth is widely resistant to the virus, based on reports of resistance from Europe. This is not the case. The virus is important because it acts biologically and can persist in the population from one season to the next, contributing to long-term population suppression. Control of the 2 nd generation must not be neglected as significant numbers of eggs can be laid throughout ust and tember and the ripening fruit is very susceptible to attack. Protection needs to be maintained throughout ust and the first half of tember. It is regrettable that no high dose sex pheromone mating disruption (MD) system is approved for use in the UK, to complement the good range of chemical control options. The Exosex autoconfusion system is the only codling moth MD system approved for use in the UK currently. Fostering the registration of other MD systems which are widely used throughout Europe (through the mutual recognition route) should be a priority for the HDC. Financial benefits The timing of insecticidal sprays against codling moth is crucial for the limitation of damage to the crop. Better timing will result in a higher yield and a more effective use of the plant protection product used. Action points for growers Do not rely solely on sex pheromone traps for codling moth monitoring. Monitor eggs at least fortnightly in orchards at risk (with a history of infestation). Agriculture and Horticulture Development Board 11. All rights reserved.

10 Use temperature based forecasting models (such as the RIMpro-Cydia model) to give supplementary information on the seasonal dynamics of the moth especially to identify periods where dusk temperature are > C. Make sure that temperature data on which predictions are based is obtained from properly calibrated met stations, ideally cited in local orchards at risk. Control measures should focus on control of the 1 st generation. They should aim to maximise the reduction of the carry-over of the populations to the following season and avoid the development of resistance.. A high standard of control needs to be maintained throughout e and y. During this period it is important to use a range of insecticides with different modes of action to minimise the risk of resistance development. Insecticides generally give 2-3 weeks protection, shorter intervals occurring in warmer weather. Use of codling moth granulovirus as part of the strategy against the 1 st generation is important. Control of the 2 nd generation must not be neglected as significant numbers of eggs can be laid throughout ust and tember and the ripening fruit is very susceptible to attack. Protection needs to be maintained throughout ust and the first half of tember. Agriculture and Horticulture Development Board 11. All rights reserved. 6

11 Science Section Introduction Objectives The seasonal dynamics of egg laying, larval development and migration to overwintering sites of codling moth through the season in relation to pheromone and kairomone trap catches of adults and meteorological conditions were studied. This was to determine the optimum timing for treatments to reduce overwintering codling moth populations to prevent high populations persisting from year-to-year. The accuracy of forecasts given by the RIMpro-Cydia model was also investigated. This will provide vital information to growers on how best to time and target sprays for codling moth against which the majority of insecticides are used on apple in the UK. Specific objectives were to: 1. determine the seasonal dynamics of egg laying by codling moth and how this relates to sex pheromone trap catches and weather conditions 2. establish the proportions of larvae completing their development at different times through the season that overwinter 3. verify the predictions of codling moth spray timing from the RIMpro-Cydia forecasting model against real biological data under UK conditions Materials and Methods Sites Two unsprayed apple orchards near Faversham and Lamberhurst, Kent which were heavily infested with codling moth were selected for the work. Site 1 In rows 1-12 of Downs organic Jonagold (M.M.6) orchard at Nichol Farm, Deerton Street, Teynham by kind permission of Nick Moor (Figure 1). Planted as Ida Red in 1978, reworked to Jonagold in 1991 (11 x 3 m tall trees/ha). Row spacing 4.4 m. Tree spacing 2.3 m. Agriculture and Horticulture Development Board 11. All rights reserved. 7

12 Site 2 In rows 8 and 14 of New Orchard Worcester Pearmain (M.26) orchard at Owl House Fruit Farm, Mount Pleasant Rd, Lamberhurst, TN3 8LY by kind permission of Colin Corfield. Planted in 199. Row spacing 3.66 m. Tree spacing in the row was 2.13 m (Figure 2). Figure 1: Location of rows at site 1 in Downs organic Jonagold orchard at Nichol Farm, Deerton Street, Teynham Agriculture and Horticulture Development Board 11. All rights reserved. 8

13 Figure 2: Location of rows at site 2 in New Worcester Pearmain orchard at Owl House Fruit Farm, Lamberhurst Figure 3: Downs organic Jonagold orchard at Nichol Farm, Deerton Street, Teynham, photographed on 1 e 9 Agriculture and Horticulture Development Board 11. All rights reserved. 9

14 Figure 4: New Worcester Pearmain orchard at Owl House Fruit Farm, Lamberhurst photographed on 1 e 9 Treatments, experimental design and layout An area (~ dwarf trees/orchard) that did not receive treatment for codling moth for the duration of the study was demarked in each orchard. Moth flight monitoring with sex pheromone and pear ester traps Two standard sex pheromone baited delta traps to monitor males and two pear ester lure (DA2313) baited delta traps to monitor males and females were deployed in each orchard in early to mid-. They were monitored weekly until ober. Lures were refreshed in early ust. The numbers of male and female codling moth in each trap were counted. Males can be distinguished from females in a number of ways including the numbers of hairs on the forewing. The females have 3 chitinised hairs (frenulum) at the shoulder just under the fore wing for holding the wing in place. The males only have one (Figure ). Males can also be deciphered by a square patch of dark coloured scales on the underside of the forewing, if the scales are still in place. This patch is absent in female codling moth. Agriculture and Horticulture Development Board 11. All rights reserved.

15 Figure : Frenulum of male tortricid moth ( Oviposition Each week, a count was made of the codling moth eggs on 3 (9) - () fruitlets per orchard. The counting was done in situ and the fruitlets were not picked. Larvae infesting fruits Codling moth damaged fruits were also collected from the 3 (9) - () fruitlets from each orchard. The larvae were removed from the fruits and the development stage of each larva was determined from the width of the head capsule. The head capsule width is a generally accepted way for determination of larval stage (see Figure 6 or frequency distribution below). There is a clear distinction in the larger larval stages, but there is some overlap for L1 and L2. To find class limits for head capsule width and duration (in Day- Degrees (DD) > C) of larval stages data from Zelger (unpublished) was used (Table 1). Agriculture and Horticulture Development Board 11. All rights reserved. 11

16 Frequency Table 1. Maximum head diameter (mm) and temperature sum (DD> C) requirements for codling moth larval stages and pupae (P) Larval Stage Head size max (mm) TSUM Cum TSUM P Head capsule diameter (mm) Figure 6: Head capsule diameter of a large sample of codling moth larvae collected in The Netherlands in 6 (H Helsen, pers. comm.) Zelger also found a relationship between the length and age of the larvae within a larval stage (unpublished). Based on Zelger s work, a frequency distribution to define four 'substages' within each of the five larval instars was used. The borders of the subclasses were set at, and 7 % of the cumulative number of larvae found in the field (Figure 7). That way a total of sub-stages ( instars x 4 sub-stages) for larval development were distinguished. For the individual larvae, the approximate date of egg deposition was backcalculated from temperature records. Agriculture and Horticulture Development Board 11. All rights reserved. 12

17 Larval length (mm) Figure 7: % development of L1-L codling moth larvae of different body lengths (L1 =dark blue; L2= pink; L3 = yellow; L4 = light blue; L = purple) Time of migration to pupation sites From mid- to ober at fortnightly intervals, replicate corrugated cardboard bands were placed around the tree trunks (covered with adhesive brown parcel tape to prevent water soaking in, Figure 8). Two weeks after deployment, the tree bands were removed and refreshed and the codling moth larvae they contained were collected and counted. The larvae collected are being held in ventilated cages in warm constant environment conditions, following a period of chilling to break diapause (outdoor cages), and the numbers of adults emerging will be determined. The data gathered will enable an estimate of the numbers of larvae pupating each month, the proportion in diapause and the proportion that are able to complete their development in the current and following season. This will provide valuable insight into the best time to focus the most intensive codling moth treatments. Agriculture and Horticulture Development Board 11. All rights reserved. 13

18 Figure 8: Corrugated cardboard bands (covered with parcel tape to make waterproof) deployed and refreshed fortnightly to monitor occurrence of larvae in pupation sites Meteorological data Weather data (half hourly readings) was taken from the official met station at East Malling Research and from the grower s weather station at Nichol Farm, Teynham. The maximum, minimum and average daily temperature (read at 9: h the following day) and the average temperature at and after dusk (average of 7 half hourly values taken between 21:-24: inclusive) were calculated. The daily temperature sum > C (a) was also calculated from the daily maximum, minimum and mean air temperatures using the triangulation method as follows If min, a = mean-. If max>.and.min< and (max+min)>, a = (max/2)+(min/4)-7. If max>.and.min< and (max+min), a= (max/4)-2. Agriculture and Horticulture Development Board 11. All rights reserved. 14

19 The weekly accumulated temperature sums > C were calculated from the daily values and forward accumulated sums for each week from the start of the codling moth flight in mid- and back accumulated sums for each week up to the end of the codling activity period in mid-ober were calculated. Accuracy of RIMpro-Cydia predictions The data was input into the RIMpro-Cydia codling moth prediction model and the accuracy of the predictions checked in relation to biological data gathered in objectives 1 and 2 as above. Comparisons were made with the standard pheromone trap and temperature sum based predictions used currently by growers. Results Moth flights Males; Sex pheromone trap catches of males (black lines on upper graphs in Figures 9-12) were similar in numbers and pattern at the two sites in any one year but the patterns of flight differed between the two years. In 9, males were captured in the first week that the traps were deployed (11-18 ) and reached a first 1 st generation peak on, then catches declined to low levels in early e with a second 1 st -generation peak being reached on 29 e, with catches falling away steadily thereafter to y. A small irregular 2 nd generation flight occurred in mid ust. In, the flight started later, was larger in total numbers and with no consistent bimodality in the 1 st generation. A small but distinct 2 nd generation flight was apparent in ust but numbers of moths captured were <% of those captured in total for the 1 st generation. A similar pattern of 1 st generation catches of males was observed in the pear ester traps but the peaks in the catches were delayed by about two weeks compared to the sex pheromone traps in 9. The delay was less obvious in. Proportionally to the numbers of males captured in the sex pheromone traps, the numbers of males captured in the pear ester traps in were smaller than in 9. Catches in the sex pheromone traps were higher in than in 9. Females: In 9, catches of females in the pear ester traps (histogram bars on upper graphs in Figures 9-12) occurred in varying numbers in most weeks between and 3 ust at site 1 (Teynham). Greatest numbers were recorded on, and and 27 Agriculture and Horticulture Development Board 11. All rights reserved.

20 y, this latter date having the greatest peak catch. Catches of females in the latter part of this period broadly coincided with the catches of males in the same traps, while the second peak of females was about two weeks later than the second peak of males. Catches of females at site 2 (Lamberhurst) were very small, only two individuals being captured on 13 y. This coincided with the peak of male catches in the pear ester traps. In, catches of females started in mid e, at least 3 weeks after the first catches of males, and continued erratically through y into early ust at both sites. No females were captured throughout the rest of ust or tember. Eggs In 9, eggs (black histogram bars on upper middle graphs in Figures 9-12) were found on fruitlets in most weeks between 22 e and 24 ust at site 1 (Teynham) with a strong first peak in the week of 6 y. At Lamberhurst, much smaller numbers of eggs were found and only between 6 and y. In, much greater numbers of eggs were found on fruitlets at Teynham. They were first observed on 9 e, rose steadily reaching a 1 st generation peak of 6 eggs/ fruits on 26 y. The 1 st generation appeared to end on ust. The 2 nd generation reached a peak on 1 tember but small numbers of eggs continued to be found until ober. Back-calculating egg laying dates from the development stages of larvae using Degree Day sums > C (white histogram bars on the upper middle graphs in Figures 9-12) indicated that in 9 at site 1 (Teynham) eggs were present in the latter part of the same period when they were directly recorded at site 1 (Teynham). At site 2 (Lamberhurst), the back-calculated eggs were present over a longer period (29 e - ust) than those seen by direct observation. In, the back-calculation of egg laying for Teynham showed more or less the same pattern as the direct egg laying observations, though no eggs were calculated to occur in the latter half or tember or early ober. Note that the back-calculation did not reveal a clear division between the 1 st and 2 nd generations. Agriculture and Horticulture Development Board 11. All rights reserved. 16

21 Larvae in fruitlets In 9, only small numbers of larvae were recorded in fruitlets between 27 y and 19 ust at site 1 (Teynham) and on 11 ust at site 2 (Lamberhurst) (white filled histogram bars on lower middle graphs in Figures 9-12). Larger numbers of larvae were recorded in. At Teynham, the first 1 st -generation larvae were found on 29 e but numbers only started to rise on the y reaching a 1 st generation peak on ust. A 2 nd generation attack started on 24 ust and continued throughout tember only ending on ober. The larval infestation at Lamberhurst in followed a broadly similar pattern, but at a much lower level. Larvae in tree bands In 9 at both sites, larvae were first found in the tree bands collected on 11/12 ust and peak numbers were found on 24 ust (grey histogram bars on bottom graphs in Figures 9 and ). Numbers gradually dwindled thereafter, but small numbers were still being found on 19 ober. Note that this was long after the crop had been picked by the farm staff, in early ober and on 3 tember at the two sites, respectively. This implies that the larvae were completing their development in the fallen fruits, or had a period of wandering, before they migrated to the tree bands to overwinter. The pattern of occurrence of larvae in tree bands followed a similar pattern in to that which occurred in 9, new larvae being found in the tree bands (which were renewed fortnightly) throughout ust and tember and into early/mid-ober. The numbers of larvae that were found in the tree bands at Teynham were surprisingly low considering the large numbers of larvae found in the fruits. Temperature and rainfall In 9, temperatures gradually rose through and e but dusk temperatures only consistently exceeded a threshold of C between 3 e and 4 y, and and 14 ust at site 1 (Teynham) and between e and 4 y, and 4 and 12 ust at site 2 (Lamberhurst) (Figures 13). It is important to note that eggs were found or were backcalculated to have occurred predominantly in the periods when air temperatures at dusk consistently exceeded C for several successive days. It was a cool summer and there were no days when dusk temperature exceeded 19 C at either site. It was also a moderately Agriculture and Horticulture Development Board 11. All rights reserved. 17

22 wet summer with significant rainfall on a high proportion of days (histogram bars in Figure 13). In, dusk temperatures (at East Malling Research, Figure 14) exceeded the threshold of C for a couple of days in the first few days of e and fairly consistently exceeded this threshold on most days from mid-e to the end of ust, with further exceedances on several days in early and late tember. In both years there was reasonably regular rainfall throughout the codling moth activity period between mid- and early ober. There was comparatively little rain in the latter 3 weeks of e in both years. Predictions of generation from accumulated temperature sums The full development of one generation of codling moth requires at least 47.7 DD>. The completion of one generation plus a 2 nd generation to L larval maturity requires 79.6 DD>. The weekly accumulated DD> values for both 9 and show total seasons accumulated DD> values from end of to end of tember of 793.9, 79.2, 722., 7.4 DD> for Teynham, nr Faversham, Kent and East Malling, Kent, respectively (Tables 2 and 3). Thus, barely enough DD> were accumulated for two generations at either site in either year, even for eggs laid in week 22 (starting Monday 9 or Monday 31 ). Individuals arising from eggs laid after week 31 in 9 (starting Monday 27 y) and after week 3 in (starting Monday 26 y) would not be able to complete their development and contribute to the following season s populations. Development from the start of pupation takes.9 DD>. Pupae that develop after week 37 in 9 (week starting Monday 7 tember ), or after week 36 in (week starting Monday 6 tember) would not have been able to complete their pupal development and emerge in the same year. Pupae found in the tree bands on 12 ust at both sites in 9 would have emerged about 3 weeks later in early tember. Those found in the bands on 24 ust would have emerged 4- weeks later. These dates are very approximate because the tree bands were only collected every two weeks and many of the individuals were larvae that had not started pupation, so the date that pupation started is imprecise. Agriculture and Horticulture Development Board 11. All rights reserved. 18

23 Table 2. Weekly temperature sums and accumulated temperature sums > C at Teynham and East Malling Research (EMR) in Kent in 9 Week (9) Teynham EMR No. Acc Acc Acc Starting Total Total Acc DD> DD> DD> to DD> to Monday: DD> DD> from start from start end end Agriculture and Horticulture Development Board 11. All rights reserved. 19

24 Table 3. Weekly temperature sums and accumulated temperature sums > C at Teynham and East Malling Research (EMR) in Kent in Week () Teynham EMR No. Acc Acc Acc Starting Total Total Acc DD> DD> DD> to DD> to Monday: DD> DD> from start from start end end Accuracy of RIMpro-Cydia predictions Predictions of the occurrence of unmated females, mated females, eggs, hatching eggs and L1 larvae were calculated by the RIMpro-Cydia model based on ½ hourly air temperature records. However, it was difficult to examine the accuracy of the model critically because it does not directly display predictions of several of the variates that were recorded in the field, which were the occurrence of males, females, eggs, larvae and late larvae/pupae. Comparisons of predicted and observed dates were thus confined to females (not Agriculture and Horticulture Development Board 11. All rights reserved.

25 distinguishing between mated and unmated) and eggs. Observed and predicted dates of the occurrence of these two variates for Teynham in 9 and are given in Table 4. The model performed well in predicting the dates of the start, middle and end of the 1 st generation females and eggs in 9. It did not appear to predict any 2 nd generation of females or eggs, though the observations do seem to indicate a small 2 nd generation did occur since eggs were recorded up to 24 ust and the sex pheromone traps indicated male flight continued well into ust and early tember, albeit at a very low level. Predictions of the timing of the 1 st generation were also reasonable in though the peak of eggs on 26 y occurred 3 weeks later than the mid-point the model predicted. However, the model failed to predict the 2 nd generation attack which was obvious and serious in the field. Table 4. RIMpro-Cydia predicted dates of 1 st, 2%, mid, 98% and end of the 1 st generation attack by codling moth and observed dates of first, mid and end the of periods of mated females and eggs at Teynham in 9 and 1st 2% Mid 98% end 9 Mated Predicted Observed Eggs Predicted Observed 22 6? Mated Predicted Observed 6 3 Eggs Predicted Observed 9 26? The start of 2 nd generation egg laying could not be distinguished form the end of the 1 st Agriculture and Horticulture Development Board 11. All rights reserved. 21

26 Discussion Seasonal dynamics of codling moth and targeting of sprays The general biology and population dynamics of codling moth are comprehensively reviewed by Van der Geest and Evenhuis (1991). The life cycle of codling moth exhibits two special features: 1) Overwintering always occurs as mature (L) diapausing larvae in a cocoon in a sheltered situation on the tree or ground; 2) The number of annual generations varies from one to four according to climate (which depends on latitude and altitude) with a proportion of larvae from each generation entering diapause. Larvae that do not complete their L development are unable to overwinter. The field observations at the two sites in the two years indicated a strong 1 st codling moth generation between mid- and end of y and a weak incomplete 2 nd generation between early ust and early ober. The temperature sum calculations indicate that only those individuals completing their 1 st generation development and laying eggs before the end of y have any chance of completing a 2 nd generation in the same year. The observations suggest that the proportion of the population that does this is zero or extremely small. Note also that 1 st generation eggs laid after the end of y will not complete their development. Thus, the 2 nd generation, though damaging to the crop, does not complete its development or contribute to the following year s population. Seasonal dynamics are strongly affected by temperatures. Not only is the development of codling moth more rapid at higher temperatures, but reproductive rates increase markedly. Hot summers greatly favour the pest. Pest monitoring methods Sex pheromone traps: These results reveal the inadequacy of relying solely on sex pheromone traps for monitoring codling moth populations and timing sprays. The sex pheromone traps gave a broad indication of the periods of codling moth activity, but could not be used to predict the occurrence of eggs in any qualitative or quantitative way. Early catches of 1 st generation males preceded the first appearance of eggs by 4- weeks in 9 but by less than two weeks in. Allowance for a delay can be made by growers, but if pheromone traps are relied on alone and a cautious approach of adopting a standard delay of two weeks is taken, this would lead to sprays being started 2-3 weeks early in some orchards/years, as in the 9 data presented here. Agriculture and Horticulture Development Board 11. All rights reserved. 22

27 More importantly, the ratio between the numbers of male moths caught in the sex pheromone traps and the numbers of eggs found on fruits in the two orchards in the two years varied greatly and was highly inconsistent. For the 1 st generation, the ratios were 93:37, 88:13, 1:237 and 138:19 for the two orchards in the two years, respectively. Furthermore, the pheromone traps scarcely indicated the 2 nd generation flights and attacks. For the 2 nd generation, the ratios of males to eggs were 4:11, 9:, 7:122 and 14:6 for the two orchards in the two years, respectively. At best only a few moths were captured in ust and tember, even at site I (Teynham) in where a heavy 2 nd generation attack occurred. The reason for this failure is unclear. Sex pheromone traps records for unsprayed orchards at the experimental stations at Randwijk, Wageningen, the Netherlands (H. Helsen, pers comm.) and at Gorsem, Belgium (T. Belien, pers comm.) in 9 and both show clear 2 nd generation moth flights (Figures 18 and 19 respectively). But based on the UK data provided by this study, if pheromone traps are to be relied on, a much lower threshold, say of two moths per trap per week, would need to be adopted for the 2 nd generation which can be very damaging as the fruit is ripening and becoming more susceptible to attack and larval penetration. Sex pheromone traps are relatively quick and easy to use and do give valuable information on the general occurrence of the pest, including of flight periods and activity, but catches need to interpreted with care and in conjunction with other monitoring and forecasting methods. If relied on alone with the existing low threshold of moths per traps per week and perhaps a lower threshold of two moths per trap per week for the 2 nd generation, excessive and not very well targeted application of sprays is likely to result. These will do an adequate job, often at higher cost than necessary, but will not optimise the reduction in the carry over of the population to the following year. Pear ester traps: The pear ester kairomone traps performed rather similarly to the sex pheromone traps in terms of capture of 1 st generation males. Numbers of males captured in the pear ester traps were much lower than captures in the sex pheromone traps, and the peaks and troughs of capture with the two trapping methods did not coincide. The start of 1 st generation pear ester traps catches were a week or two later than the sex pheromone trap catches, probably due to reduced attractiveness. The pear ester traps did not give any indication of the 2 nd generation (of males or females) and this was a serious failing. The main advantage of the pear ester traps is that they capture females, though numbers captured in this study were small and erratic, and no 2 nd generation females were caught at Agriculture and Horticulture Development Board 11. All rights reserved. 23

28 all. The catches of females did roughly coincide with the periods of egg-laying in broad terms, but the erratic nature and small size of the catches could not be relied on to give a clear indication of the times of egg laying or of the numbers of eggs laid. Again there was no consistent ratio between the numbers of females captured and the numbers of eggs laid. The moths are easy to sex by the presence of a characteristic dark patch on the undersides of the forewing of males which is absent in the females. We did not attempt to determine whether or not he females were mated, which can be done by dissection and examination for spermatophores under a binocular microscope in the laboratory. The failure to catch 2 nd generation females was probably due to lack of attractiveness of the pear ester lures and increasing competition with the volatiles that were produced by the fruits as they ripened. The use of the pear ester traps is worthy of further investigation but this study did not indicate their use would make a substantial contribution to improving codling moth monitoring. Egg monitoring: Regular, direct counting of eggs on fruitlets in situ is probably the only fully reliable and accurate way of determining the timing and need for sprays against codling moth. Direct egg monitoring is time consuming and requires diligence and good eye sight, though is not difficult. Observations would need to be done weekly, at least till the first insecticide spray is applied, but the frequency of monitoring could possibly be reduced to fortnightly (the approximate persistence of insecticides in warm weather) at times when a protective insecticide spray deposit is present. The sample size needs to be large as there is a low economic threshold for codling moth damage as fruits are attacked directly (<.% of fruits). Ideally at least 6 fruitlets per orchard, from throughout the orchard, would be examined on each occasion. Probability statistics indicates that a sample size of 98 fruitlets would be needed to detect an infestation of.% fruits with 9% confidence, providing the efficacy of the examination is % (Table ). The sample size needed increases sharply as the efficacy of the sampler decreases, as the confidence level required increases and as the % level of detection decreases (Table ). However, stratifying the sampling (targeting it to where codling moth damage is most likely to occur on fruits exposed to sunshine at the top of the tree) will markedly improve the efficiency of sampling. Furthermore, when egg numbers are high the sampling size could be reduced. The time required to examine 6 fruitlets is at least an hour, so such sampling is costly. However, codling moth control is probably costing growers at least 3 per ha per season and better targeting and timing of sprays might reduce this by at least per ha per season, so some investment in egg monitoring is justified. Agriculture and Horticulture Development Board 11. All rights reserved. 24

29 Table. Sample sizes for 9% and 99% confidence at varying levels of detection and efficacies of detection (lot size large and mixed, binomial distribution) % efficacy of sampler P = 9% (confidence level) % level of detection P = 99% (confidence level) % level of detection Forecasting models In view of the limited value of the information provided by sex pheromone trap monitoring, computer based phenological simulation models are an important source of information for interpreting codling moth seasonal dynamics. The RIMpro-Cydia model is the most sophisticated model available. It did give broadly correct predictions of the 1 st generation seasonal dynamics, taking into account well established day-degree air temperature sums > C as well as dusk temperatures > C which are known to be important for egg laying activity. However, the RIMpro-Cydia model failed to predict the occurrence of the 2 nd generation attack in both years, a particularly serious failing at Teynham in where large numbers of eggs were laid throughout ust and into tember which resulted in a significant larval attack. Temperature based models are sensitive to air temperatures and the accuracy of grower operated weather stations needs to be scrutinised. Note that the East Malling Met station is a high quality one calibrated by the Met office. The RIMpro-Cydia model did predict the occurrence of a 2 nd generation flight based on temperature data from Tonbridge (Figure 17). Agriculture and Horticulture Development Board 11. All rights reserved.

30 Control tactics The seasonal dynamics of codling moth has implications for control tactics: It implies that control measures to maximise the reduction of the carry over of the populations to the following season and avoid the development of resistance should focus on control of the 1 st generation. The records indicate that egg-laying does not start till a week or two after the first catches of moths in sex pheromone traps. A high standard of control thus needs to be maintained throughout e and y. During this period it is important to use a range of selective insecticides with different modes of action to minimise the risk of resistance development. A list of the insecticide and biological products approved for use on apple and pear and their class, modes of action and limitations on use is given in Table 6. Table 6. Pesticides approved for use on apple and pear in the UK which may be used for control of codling moth Harves Max no. Example IRAC t Active substance Class per product mode internal season (days) Bacillus thuringiensis Dipel Microbial toxin - U chlorantraniliprole Coragen Anthranilic chlorpyrifos Dursban OP 1B 14 3 cypermethrin Toppel SP 3 codlemone Exosex Sex phero MD - 2 codling moth granulovirus Cyd-X Microbial biocontrol deltamethrin Decis SP 3 U diflubenzuron Dimilin CSI 14 3 fenoxycarb Insegar JHA 7B 42 2 indoxacarb Steward Oxadiazine 22A 7 3 methoxyfenozide Runner MAC 18A 14 3 spinosad Tracer Neural blocker 7 4 thiacloprid Calypso Neoniotinoid 4A 14 3 U = unlimited Not specifically approved/recommended for codling moth control Bt has only weak activity against codling moth Agriculture and Horticulture Development Board 11. All rights reserved. 26

31 Use of codling moth granulovirus as part of the strategy against the 1 st generation is also important. Some UK growers may be under the misapprehension that codling moth is widely resistant to the virus, based on reports of resistance from Europe. This is not the case. Resistance only occurred in a few organic orchards in northern Europe where the virus was used intensively over many years, and in any case the strain of the virus has now been altered to overcome the problem. The virus is important because it acts biologically and can persist in the population from one season to the next, contributing to long term population suppression. Control of the 2 nd generation must not be neglected as significant numbers of eggs can be laid throughout ust and tember. Protection needs to be maintained throughout ust and the first half of tember. It is regrettable that no high dose sex pheromone mating disruption (MD) system is approved for use in the UK, to complement the good range of. The Exosex autoconfusion system is the only codling moth MD system approved for use in the UK currently. Fostering the registration of other MD systems which are widely used throughout Europe (though the mutual recognition route) would be invaluable and should be a priority for the HDC. Conclusions Seasonal dynamics of codling moth and targeting of sprays o o o There was a strong 1 st codling moth generation between mid- and end of y and a weak incomplete 2 nd generation between early ust and tember or early ober at both sites in both years Temperature sum calculations indicate that only those individuals completing their 1 st generation development and laying eggs before the end of y have any chance of completing a 2 nd generation in the same year. In 9 and the proportion of the population that did this was zero 1 st generation eggs laid after the end of y did not complete their development. Thus, the 2 nd generation, though damaging to the crop, does not complete its development or contribute to the following year s population Agriculture and Horticulture Development Board 11. All rights reserved. 27

32 Sex pheromone traps o This study confirms that sex pheromone traps should not be relied on solely for monitoring codling moth populations and timing sprays. They gave a broad indication of the periods of codling moth activity, but could not be used to predict the occurrence of eggs in any qualitative or quantitative way o Early catches of 1 st generation males preceded the first appearance of eggs by 4- weeks in 9 but by less than two weeks in o Importantly, the ratio between the numbers of male moths caught in the sex pheromone traps and the numbers of eggs found on fruits in the two orchards over the two years varied greatly and was highly inconsistent. For the 1 st generation, the ratios were 93:37, 88:13, 1:237 and 138:19 for the two orchards in the two years, respectively. Furthermore, the pheromone traps scarcely indicated the 2 nd generation flights and attacks. For the 2nd generation, the ratios of males to eggs were 4:11, 9:, 7:122 and 14:6 for the two orchards over the two years, respectively. At best only a few moths were captured in ust and tember, even at one site (Teynham) in where a heavy 2 nd generation attack occurred o Based on the UK data provided by this study, if pheromone traps were to be relied on, a much lower threshold, say of two moths per trap per week, would need to be adopted for the 2 nd generation which can be very damaging as the fruit is ripening and becoming more susceptible to attack and larval penetration o Sex pheromone traps are relatively quick and easy to use and do give valuable information on the general occurrence of the pest, including of flight periods and activity, but catches need to be interpreted with care and in conjunction with other monitoring and forecasting methods. If relied on alone with the existing low threshold of five moths per trap per week and perhaps a lower threshold of two moths per trap per week for the 2 nd generation, excessive and not very well targeted application of sprays is likely to result, which will do an adequate job but often at higher cost than necessary. In addition, the reduction in the carry over of the population to the following year will not be optimised Pear ester traps o The pear ester kairomone traps performed rather similarly to the sex pheromone traps in terms of capture of 1 st generation males, though numbers of males captured in the pear ester traps were much lower than captures in the sex pheromone traps; also the peaks and troughs of capture with the two trapping methods did not coincide. The start of 1 st generation pear ester traps catches were a week or two later than the Agriculture and Horticulture Development Board 11. All rights reserved. 28

33 sex pheromone trap catches, probably due to reduced attractiveness. The pear ester traps did not give any indication of the 2 nd generation (of males or females) and this was a serious failing o The main advantage of the pear ester traps is that they capture females, though numbers captured in this study were small and erratic, and no 2 nd generation females were caught at all. The catches of females did roughly coincide with the periods of egg-laying, but the erratic nature and small size of the catches could not be relied on to give a clear indication of the times of egg laying or of the numbers of eggs. Again there was no consistent ratio between the numbers of females captured and the numbers of eggs. The moths are easy to sex by the presence of a characteristic dark patch on the undersides of the forewing of males which is absent in the females o The failure to catch 2 nd generation females was probably due to a lack of attractiveness of the pear ester lures and increasing competition with the volatiles that were produced by the fruits as they ripened. The use of the pear ester traps is worthy of further investigation but this study did not indicate their use would make a substantial contribution to improving codling moth monitoring Egg monitoring o o o o Regular, direct counting of eggs on fruitlets in situ is probably the only fully reliable and accurate way of determining the timing and need for sprays against codling moth currently Direct egg monitoring is time consuming and requires diligence and good eye sight, but is not difficult Observations ideally would need to be done weekly with a sample size of ideally at least 6 fruitlets per orchard per sampling occasion. Targeting the sampling to where codling moth damage is most likely to occur on fruits exposed to sunshine at the top of the tree will markedly improve efficiency. The sample size could be reduced when egg numbers are high The time required to examine 6 fruitlets is at least an hour, so such sampling is costly. However, codling moth control is probably costing growers at least 3 per ha per season and better targeting and timing of sprays might reduce this by at least per ha per season, so investment in egg monitoring is justified Agriculture and Horticulture Development Board 11. All rights reserved. 29

34 Forecasting models o o o In view of the limited value of the information provided by sex pheromone trap monitoring, computer based phenological simulation models are an important source of information for interpreting the seasonal dynamics of codling moth The RIMpro-Cydia model is the most sophisticated model available. It did give broadly correct predictions of the 1 st generation seasonal dynamics, taking into account well established day-degree air temperature sums > C as well as dusk temperatures > C which are known to be important for egg laying activity However, the RIMpro-Cydia model failed to predict the occurrence of the 2 nd generation attack in both years, a particularly serious failing at Teynham in where large numbers of eggs were laid throughout ust and into tember which resulted in a significant larval attack. Temperature based models are sensitive to air temperatures and the accuracy of grower operated weather stations needs to be scrutinised Control tactics o o o o o The seasonal dynamics of codling moth in the UK has implications for control tactics. Control measures to maximise the reduction of the carry over of the populations to the following season and avoid the development of resistance should focus on control of the 1 st generation A high standard of control thus needs to be maintained throughout e and y. During this period it is important to use a range of insecticides with different modes of action to minimise the risk of resistance development Use of codling moth granulovirus as part of the strategy against the 1 st generation is important. Some UK growers may be under the misapprehension that codling moth is widely resistant to the virus, based on reports of resistance from Europe. This is not the case. The virus is important because it acts biologically and can persist in the population from one season to the next, contributing to long-term population suppression Control of the 2 nd generation must not be neglected as significant numbers of eggs can be laid throughout ust and tember and the ripening fruit is very susceptible to attack. Protection needs to be maintained throughout ust and the first half of tember It is regrettable that no high dose sex pheromone mating disruption (MD) system is approved for use in the UK, to complement the good range of chemical control options. The Exosex autoconfusion system is the only codling moth MD system Agriculture and Horticulture Development Board 11. All rights reserved. 3

35 approved for use in the UK currently. Fostering the registration of other MD systems which are widely used throughout Europe (though the mutual recognition route) should be a priority for the HDC Agriculture and Horticulture Development Board 11. All rights reserved. 31

36 Mean no. / bands Mean no. / week Mean no. / week Mean no. males / week Mean no. females / week Moths / trap Ester females 4 Pheromone males 8 3 Ester males Eggs / 3 fruits 4 3 Observed on fruits Back-calculated from larvae Larvae / 3 fruits Larvae / trunk bands Figure 9: Site 1 9, Nichol Farm, Teynham. Top) Catches of adult moths in pheromone and pear ester traps; Upper middle) Eggs on fruits; Lower Middle) Larvae in fruits; Bottom) Larvae in trunk bands. Agriculture and Horticulture Development Board 11. All rights reserved. 32

37 Mean no. / week Mean no. / week Mean no. males / week mean no. females / week Moths / trap Ester females Eggs / 3 fruits Observed on fruits Larvae / 3 fruits Agriculture and Horticulture Development Board 11. All rights reserved. 33

38 Mean no. / week Mean no. males / week Mean no. females / week Mean no. / bands Larvae / trunk bands Figure : Site 2 9, Owl House Farm, Lamberhurst. Top) Catches of adult moths in pheromone and pear ester traps; Upper middle) Eggs on fruits; Lower Middle) Larvae in fruits; Bottom) Larvae in trunk bands Moths / trap Ester females Eggs / fruits Observed on fruits Agriculture and Horticulture Development Board 11. All rights reserved. 34

39 Mean no. males / week Mean no. females / week Mean no. / bands Mean no. / fruits Larvae / fruits Larvae / bands Figure 11: Site 1, Nichol Farm, Teynham. Top) Catches of adult moths in pheromone and pear ester traps; Upper middle) Eggs on fruits; Lower Middle) Larvae in fruits; Bottom) Larvae in trunk bands. Moths / trap Ester females Agriculture and Horticulture Development Board 11. All rights reserved. 3

40 Mean no. / bands Mean no. / fruits Mean no. / fruits Eggs / fruits 6 Observed on fruits Back-calculated from larvae Larvae / fruits Larvae / bands Figure 12: Site 2, Owl House Farm, Lamberhurst. Top) Catches of adult moths in pheromone and pear ester traps; Upper middle) Eggs on fruits; Lower Middle) Larvae in fruits; Bottom) Larvae in trunk bands Agriculture and Horticulture Development Board 11. All rights reserved. 36

41 Temp (C) Rainfall (mm) Temp (C) Rainfall (mm) Weather data "Teynham" Kent Rainfall (mm) EMR Max temp C Min temp C Dusk temp C (21: - 24:) 3 Weather data "Brenchley" Kent Rainfall (mm) EMR Max temp C Min temp C Dusk temp C (21: - 24:) Figure 13. Met data from Teynham and Brenchley Kent in 9 Agriculture and Horticulture Development Board 11. All rights reserved. 37

42 Temp ( C) Rainfall (mm) Temp ( C) Rainfall (mm) Weather data for "Nicol farm, Teynham" Kent Rainfall (mm) Min Temp Max Temp Weather data for "East Malling Research" Kent Rain Fall Max Temp (ºC) Date Dusk Temp (ºC) Min Temp (ºC) Figure 14. Met data for Agriculture and Horticulture Development Board 11. All rights reserved. 38

43 Figure. Predictions of the occurrence of unmated females (light blue), mated females (dark blue), eggs (yellow) and hatching larvae (red) provided by the RIMpro- Cydia model based on ½ hourly air temperature records taken at Nichol Farm Teynham in Kent (Site 1) in 9. The predictions of eggs and hatching larvae corresponded well in general with actual records Figure 16. Predictions of the occurrence of unmated females (light blue), mated females (dark blue), eggs (yellow) and hatching larvae (red) provided by the RIMpro- Agriculture and Horticulture Development Board 11. All rights reserved. 39

44 Cydia model based on ½ hourly air temperature records taken at Nichol Farm Teynham in Kent (Site 1) in Figure 17. Predictions of the occurrence of unmated females (light blue), mated females (dark blue), eggs (yellow) and hatching larvae (red) provided by the RIMpro-Cydia model based on ½ hourly air temperature records taken at Moat Farm Tonbridge in Kent in (Supplied by T Biddlecombe) Agriculture and Horticulture Development Board 11. All rights reserved. 4