Survival of Pseudomonas syringae pv. actinidiae P. s. pv. syringae

Survival of Pseudomonas syringae pv. actinidiae and P. s. pv. syringae bacteria on honey bees and in beehives

Pattemore DE, Hoyte SM, McBrydie HM, Goodwin RM, Moffat B, Yu J, Parry F, and Ah Chee A

October 2011

A report prepared for

ZESPRI Group Ltd V11255

Pattemore DE, Hoyte SM, McBrydie HM, Goodwin RM, Moffat B, Yu J, Parry F and Ah Chee A.

Plant & Food Research, Ruakura

SPTS No. 6109

DISCLAIMER

Unless agreed otherwise, The New Zealand Institute for Plant & Food Research Limited does not give any prediction, warranty or assurance in relation to the accuracy of or fitness for any particular use or application of, any information or scientific or other result contained in this report. Neither Plant & Food Research nor any of its employees shall be liable for any cost (including legal costs), claim, liability, loss, damage, injury or the like, which may be suffered or incurred as a direct or indirect result of the reliance by any person on any information contained in this report.

LIMITED PROTECTION

This report may be reproduced in full, but not in part, without prior consent of the author or of the Chief Executive Officer, The New Zealand Institute for Plant & Food Research Ltd, Private Bag 92169, Victoria Street West, Auckland 1142, New Zealand.

CONFIDENTIALITY

This report contains valuable information in relation to the ZESPRI Psa research programme that is confidential to the business of Plant & Food Research and ZESPRI Group Limited. This report is provided solely for the purpose of advising on the progress of the ZESPRI Psa research programme, and the information it contains should be treated as "Confidential Information" in accordance with the Plant & Food Research Agreement with ZESPRI Group Limited.

This report has been prepared by The New Zealand Institute for Plant & Food Research Limited (Plant & Food Research), which has its Head Office at 120 Mt Albert Rd, Mt Albert, Auckland.

This report has been approved by:

David Pattemore

Scientist, Pollination & Apiculture

Date: 25 October 2011

Bob Fullerton

Science Group Leader, Pathology and Applied Mycology

Date: 25 October 2011

Contents

Executive summary i

1 Introduction 1 1.1 Background 1 1.2 Aims and Study Design 2

2 Methods 3 2.1 Temperature and Humidity Records 3 2.2 Selection of Psa analogue 3 2.3 Survival and transmission of Pss_SMR in beehives 4 2.4 Survival of Psa in beehives 5 2.5 Survival of Psa on kiwifruit pollen at 35°C 5 2.6 Correlation between Pss_SMR survival and pollen-grain retention 6

3 Results 8 3.1 Temperature Comparison 8 3.2 Selection of Psa analogue 9 3.3 Survival and transmission of Pss_SMR in beehives 10 3.4 Survival of Psa in beehives 13 3.5 Survival of Psa on kiwifruit pollen at 35°C 14 3.6 Correlation between Pss_SMR survival and pollen-grain retention 15

4 Discussion 18 4.1 In conclusion 18

5 Recommendations 19 5.1 Honey Bee Management Plan 19 5.2 Possible Legislative Tools 20

6 Acknowledgements 20

7 References 20

The New Zealand Institute for Plant & Food Research Limited (2011) Page i This report is confidential to Plant & Food Research and ZESPRI Group Limited Survival of Pseudomonas syringae pv. actinidiae and P. s. pv. syringae on bees and in beehives. SPTS No. 6109

Executive summary

Survival of Pseudomonas syringae pv. actinidiae and P. s. pv. syringae on bees and in beehives

Pattemore DE, Hoyte SM, McBrydie HM, Goodwin RM, Moffat B, Yu J, Parry F, Ah Chee A.

October 2011, SPTS No. 6109

Pseudomonas syringae pv. actinidiae (Psa) is a serious disease of kiwifruit (Actinidia spp.), and was first confirmed in New Zealand in November 2010. Honey bees collect pollen from flowers and spread this to different vines, but it is not known how important bees may be in spreading Psa via contaminated pollen. The aims of this project were to 1) assess whether Psa could be spread from one group of foraging bees to another group in the same hive that were foraging in a different location, and 2) assess the length of time Psa can survive on bees and in beehives, in order to assess the risk of spreading the disease by the movement of hives.

To investigate whether Psa can be spread within a hive, pollen contaminated with streptomycin-resistant Pseudomonas syringae pv. syringae (Pss_SMR) was presented to marked pollen-foraging bees, and was also used to cover a sample of marked bees from beehives. Pss_SMR was recovered from marked bees 14 days after first coming in contact with contaminated pollen, and was also recovered from unmarked bees on the day that the bees were presented with contaminated pollen and for up to six days afterwards.

To assess the length of time Psa can survive in beehives and on bees, live caged bees and plastic discs were inoculated with Psa (in containment) and Pss_SMR (in the field) and were placed inside beehives. Pss_SMR and Psa were shown to survive for nine days on plastic discs in ambient laboratory conditions, but while Pss_SMR was still found on discs after 14 days in beehives in the field, Psa on discs was only found for two days within hives in the containment facility. In comparison, both Psa and Pss_SMR were found on caged bees after six days but not on subsequent sample days.

These results show that Pseudomonas spp. bacteria on pollen can be spread rapidly within a hive, so that bees from a single hive may be able to spread Psa between orchards without the hive being moved. These bacteria can survive for two weeks or more on bees or on surfaces within the hive, so bees may be able to spread Psa between orchards. The rapid decline of bacterial populations in the first 24 h means that few bacteria may remain for long periods of time but, as it is still unknown how many bacteria are required to start an infection, it is prudent to consider hives as potential sources of Psa for at least 14 days.

For further information please contact:

David Pattemore The New Zealand Institute for Plant & Food Research Ltd Plant & Food Research, Ruakura Private Bag 3230, Waikato Mail Centre Hamilton 3240 NEW ZEALAND Tel: +64-7-959 4459 Fax: +64-7-959 4431 Email: [email protected]

The New Zealand Institute for Plant & Food Research Limited (2011) Page ii This report is confidential to Plant & Food Research and ZESPRI Group Limited Survival of Pseudomonas syringae pv. actinidiae and P. s. pv. syringae on bees and in beehives. SPTS No. 6109

The New Zealand Institute for Plant & Food Research Limited (2011) Page 1 This report is confidential to Plant & Food Research and ZESPRI Group Limited Survival of Pseudomonas syringae pv. actinidiae and P. s. pv. syringae on bees and in beehives. SPTS No. 6109

1 Introduction

1.1 Background

In response to the spread of the virulent bacterial disease Pseudomonas syringae pv. actinidiae (Psa) in kiwifruit orchards in New Zealand, the New Zealand Institute for Plant & Food Research Limited (PFR) has established a dedicated Psa research programme. A key outcome within the on-orchard management research area is to develop “Psa-free pollination systems to enable growers to make informed decisions on pollination in their orchards”.

This report has been commissioned by and prepared for ZESPRI Group Ltd as part of their response to the spread of Psa.

Although preliminary testing of bees and beehives did not detect the presence of Psa (Shane Max, ZESPRI, pers. comm.), positive results from PCR tests on kiwifruit pollen indicate that contamination of pollen is indeed possible (Joel Vanneste, PFR, pers. comm.), and thus would probably contaminate foraging honey bees and the hives from which they come.

There are four main ways that honey bees could spread Psa bacteria between orchards:

1. On the outside of beehives, trucks and beekeeping equipment that is moved between orchards

2. Beehives that are moved out of orchards and are placed in farmland that is within flight distance of other orchards that are still in flower

3. Bees that have been foraging on kiwifruit flowers in an infected orchard and are then moved into a second orchard and start foraging on new kiwifruit flowers

4. Honey bees that become contaminated in one orchard and on return to their hive, contaminate other bees that are foraging in a second orchard.

The risk of carrying Psa on the outside of beehives and on beekeeping equipment can best be reduced by controlling hive movements and sterilizing equipment.

The potential problem presented by hives no longer being used for pollination but being situated close to other orchards that are still in flower can be managed with movement restrictions.

The spread of Psa by foraging bees from hives that are moved between orchards can be managed by specifying the minimum time hives must be held between removing from one orchard and being moved to a second orchard, as well as placing restrictions on the areas between which bees can be moved (as above). As yet, we do not know the length of time hives should be held between being removed from an infected orchard and being placed in a second orchard to minimise the risk of spreading Psa via this pathway.

The possibility that bees foraging from a hive in an infected orchard may contaminate bees foraging from the hive in a second orchard cannot be managed, and this would have implications for the repeated use of hives within an infected zone.


1.2 Aims and Study Design

The aim of this project was to determine the length of time that Psa bacteria are likely to survive in a beehive. A secondary aim was to assess the degree to which bees visiting an infected orchard could contaminate bees from the same hive that are foraging in another location.

The ability of Psa to survive within a hive will be dependent on its ability to survive on bees and hive materials at the temperatures and humidity levels commonly found within hives. This study did not address whether surviving Psa within beehives can then be spread to and infect kiwifruit vines.

Because of the strict controls on the use of Psa in experiments, we were only able to conduct experiments on the survival of Psa in Physical Containment Level 1+ (PC1+) facilities. As these containment laboratory conditions do not accurately represent conditions that Psa would encounter in orchards, a related Pseudomonas bacteria was used in field experiments to assess the survival of plant pathogenic bacterium in realistic hive conditions.

Accordingly, the project had three main stages:

1. Identify and culture an antibiotic-resistant isolate of Pseudomonas syringae pv. syringae (Pss) to use as an analogue of Psa

2. Assess survival of Psa in beehives on inoculated bees and plastic discs in a PC1+ containment facility

3. Assess survival of Pss in beehives in the field, on a) bees foraging on Pss-contaminated pollen and on b) inoculated bees and plastic discs.

Two additional studies were conducted following initial analysis of results. The aims of these studies were to:

1. Assess the survival of Psa on pollen and on plastic discs in an incubator kept at 35°C.

2. Investigate the correlation between survival of bacteria and retention of kiwifruit pollen-grains on marked honey bees.


2 Methods

2.1 Temperature and Humidity Records

To compare the growing environments for bacteria between laboratory, beehives in the field and beehives in the containment facility, probes were used to record temperature (°C) and relative humidity (RH) during the course of the experiment. Probes in beehives were placed in the outer edges of hives and in the centre of brood. Data were recording hourly or half-hourly for the duration of each experiment, and the resulting data were plotted for comparison. The significance of differences in RH recorded in the different locations were tested using two-tailed, independent-sample t-tests at α = 0.05, and using a conservative Bonferroni correction for multiple comparisons (reducing the required P-value for significance to 0.005).

2.2 Selection of Psa analogue

For this project an appropriate antibiotic-resistant bacterial strain developed to function as an analogue for Psa, was used to test methods for contaminating pollen, and test bees for its survival.

2.2.1 Generation of an antibiotic-resistant strain of Pss

Initially, a rifampicin and streptomycin-resistant strain of Pss was produced using the method of Walker et al. (2004). In summary, a wild-type culture of Pss was cultured in tryptone soya broth (TSB) and transferred to nutrient agar (NA) containing 10 µg/ml streptomycin. The resultant streptomycin-resistant colonies were then streak plated onto NA containing 100 µg/ml rifampicin (NAR). The rif+ marked strains were maintained on this medium. This initial strain grew slowly and was too weak to be used in trials. Therefore, an alternative streptomycin-resistant strain (Pss_SMR) was produced using the above methods.

2.2.2 Combining pollen and viable bacterial cells

Several drying techniques for bacteria were evaluated and a method identified which minimised bacterial mortality. Dried Pss_SMR bacterial cells and pollen were combined and the concentrations of viable bacteria (colony forming units, CFU count) were determined using dilution plating techniques on NA.

2.2.3 Validation of the recovery of Pss_SMR from bees

Pollen artificially inoculated with Pss_SMR was applied to bees and several washing and recovery techniques were evaluated. These included comparisons of rates of wetting agent and plus and minus sonification to optimise recovery rates on NA.

2.2.4 Growth comparison to Psa

Plastic discs were inoculated with 10 µL of Pss_SMR or Psa broth, and then stored on a tray inside a plastic bag at room temperature in a PC2 laboratory for nine days. Three discs each of Pss_SMR and Psa were then sampled at each of the following time points (days): 0, 1, 2, 5, 8 and 9. Sampled discs were washed in 1 mL of sterile distilled water (SDW) and serially diluted 5 times before being plated in 10-µl droplets with three replicates onto King’s B agar (KB) in the case of Psa or streptomycin KB in the case of Pss_SMR). After 48 h of incubation at 28°C, CFU counts were recorded. The survival of Pss_SMR and Psa over the nine days was compared.


2.3 Survival and transmission of Pss_SMR in beehives

2.3.1 Pss_SMR survival and transmission by pollen foraging bees

Sixty grams of artificially collected kiwifruit pollen was inoculated with Pss_SMR. To achieve this, the Pss_SMR isolate was cultured on 120 Petri dishes of streptomycin KB for 3 days at 20°C. The bacteria from two Petri dishes of culture were removed by adding 4 mL of SDW to each and scraping gently with a flame-sterilised glass hockey stick. The bacterial suspension was collected in a sterile 50-mL Falcon™ tube, made up to 20 mL with SDW and mixed on a vortex stirrer. Two grams of kiwifruit pollen was weighed out in each of 30 similar 50-mL tubes. The 20 mL of bacterial suspension was added directly to the tubes containing the pollen and these were mixed on the vortex stirrer for 30 sec, allowed to sit for 1-2 min and then mixed again for 10 sec. This mixture of bacteria and pollen was spread thinly onto plastic trays (450 mm x 310 mm) and dried in a lamina flow hood for 3-4 h with the room temperature set to 25°C to aid drying. The dried mixture was carefully scraped off the plastic trays and collected in sterile 50-mL tubes. The dried contaminated pollen was combined and gently milled in a mortar and pestle for ~10 sec to create a flow-able powder ready for feeding to the bees.

Bees from six hives were trained to collect kiwifruit pollen at a feeding station outside their hive, and 12 bees were captured as background controls before the contaminated pollen was presented. After the contaminated pollen was placed in the feeding station, foraging bees were captured, anoxiated with carbon dioxide (CO2), marked with white acetone-based paint, and then released to return to their hives. Twenty bees that were captured feeding on pollen were taken as time zero samples, and ten of these were processed after their pollen sac had been removed.

Marked bees were then captured from hives on days 1, 2, 5, 9 and 13, and washed to quantify the number of Pss_SMR colony forming units (CFU). Individual bees were washed in 1 mL of SDW (amended with 0.01% Tween®80) by vortexing for 30 sec and making serial dilutions.

Appropriate dilutions had 20-L droplets dried onto Petri dishes containing streptomycin KB. Colony forming units were counted after 48 h of incubation at 20°C.

On the day the contaminated pollen was presented to bees, other pollen-foraging bees returning to the hives with different pollen (identifiable by colour) were sampled to assess the degree of contamination within hives between different groups of pollen foragers (five bees from each of two hives). To compare the mean number of CFU/mL between groups of bees from day 0, we conducted pair-wise t-tests (using Bonferonni’s correction to derive a significance level of 0.0167 from an initial desired α = 0.05).

Initial counts found 54 marked bees in one hive, 15 in another, and just one marked bee in a third, but no marked bees were found in the remaining three hives. Because of this uneven distribution of marked bees and low recapture rate, a slightly different experiment was set up to provide better replication of the transmission of contaminated material between marked and unmarked bees.

For the second experiment, 100 bees were caught from each of five hives (different hives from those originally used, which were located 1.4 km away), anoxiated with CO2, and marked with white paint. The marked bees were then coated with Pss_SMR contaminated pollen (~5 g per 100 bees), and released in front of their hives. Two unmarked, uncontaminated bees from each hive were sampled as background controls, and two marked and contaminated bees from each lot of 100 bees were sampled on the day of contamination. Five marked and five unmarked


bees were then sampled from each of the five hives on days 1, 2, 5, 9 and 19 and washed to quantify the number of Pss_SMR colony forming units as described above.

2.3.2 Pss_SMR survival in beehives

At the same time as the experiment above, 55 bees from each hive (located on the Ruakura Research Station) were captured, anoxiated with CO2, and then inoculated with a 10-µL drop of Pss_SMR broth. Inoculated bees were placed in 75 mm x 33 mm x 15 mm plastic cages (five bees in each of ten cages per hive). Five of the cages were then placed back into each beehive on the outer edge, and five were placed in the centre of the brood. Five inoculated bees from each hive were used as time zero samples. At the same time, 1-cm diameter PVC plastic discs were also inoculated with a 10-µL drop of Pss_SMR broth. Five inoculated discs were then secured in a plastic cage (two per hive) so that they were not touching. One cage of discs was placed in the centre of the brood, and the other at the outer edge of each hive. A further cage with five control (not inoculated) discs was also placed at the outer edge of each hive.

Sampling was conducted on days 1, 2, 6, 9 and 14. On each sampling day, one cage of bees and one inoculated disc from the brood, and one cage of bees, one inoculated disc and one control disc from the outer edge of each hive were sampled. Each cage of five bees were chilled for approximately two minutes at -80°C, and then placed into 2.5 mL of sterile distilled water. Discs were placed into 1 mL of sterile distilled water. Samples were then vortexed for 30 sec and the number of Pss_SMR CFU were determined as described above.

2.4 Survival of Psa in beehives

In addition to the Pss_SMR survival trial in Section 2.3 ii) above, a similar methodology was used to assess the survival of Psa in beehives. Because of the constraints related to the use of Psa, beehives for these trials were kept inside an insect-proof two-room tent positioned inside a secure PC1+ containment facility (Ministry of Agriculture and Forestry registration number 2744). As space was limited, only two hives were used for this experiment, and each hive contained approximately 20,000 bees. The bees were fed pollen substitute, sugar water and water. After the experiment was concluded, the hives were killed and the tent, hives and all the equipment were disposed in biohazard waste. It is important to note that the activity and behaviour of the bees in hives inside the tent in this containment facility will be very different from wild bees in the field, and therefore the results may not be indicative of what could be expected in orchards this spring.

Cages of inoculated bees and discs were placed inside each of the two hives as per the methods described above for PSS_SMR in Section 2.3 ii), and sampled on days 1, 2, 6, 9 and 14.

2.5 Survival of Psa on kiwifruit pollen at 35°C

To test the survival of Psa at the temperature experienced at the centre of brood in a honey bee hive (35°C), plastic discs were inoculated in the same manner as in section 2.2.4 and three different samples of male kiwifruit pollen were contaminated with Psa in same manner as that used for Pss_SMR in section 2.3.1

Equal numbers of plastic discs were placed on trays inside plastic bags on a laboratory bench and in an incubator at 35°C. Threes discs were removed at each time point (days 0,1,2,4,8),


washed in 1 mL of SDW and serially diluted 5 times before being plated in 10-µl droplets with three replicates onto KB.

The three samples of contaminated pollen were each divided into two containers, with one set of pollen samples kept on a laboratory bench and one set in the 35°C incubator. At each sample time point (days 0,1,2,4,8), approximately 0.0027g of contaminated pollen was taken from each sample and placed in 1mL of SDW and serially diluted 5-7 times before being plated in 10-µl droplets with three replicates onto KB.

We compared the survival of Psa on pollen at 35°C and ambient laboratory temperature, and compared this to survival on plastic discs.

2.6 Correlation between Pss_SMR survival and pollen-grain retention

To assess whether the length of time that Pss_SMR survived on pollen-coated and pollen-collecting bees (section 2.3.1) could be explained by the retention of Pss_SMR-contaminated pollen on the bees, the stock solutions sampled from marked and unmarked bees were re-sampled to assess the concentration of pollen grains.

For this study, a sub-sample of the bees was assessed for pollen counts. From the first experiment involving the presentation of contaminated pollen to pollen-foraging bees, pollen grains were counted for all marked bees from Hives 1 and 2 that were sampled on days 1 and 2, as well as five randomly selected bees from day 0 with pollen sacks and five bees from day 0 with pollen sacks removed. Bees from day 0 that had their pollen sacks removed before sampling were treated as day 0.5 samples in the analysis, as they were analogous to bees that had returned to the hive and removed their pollen load.

From the second experiment where 100 bees from each hive were coated with contaminated pollen, two marked bees from each hive from days 1, 2 and 5 were sampled for pollen, along with at least one unmarked bee from each hive on each day and any unmarked bees that returned positive results for Pss_SMR and up to two marked bees from each hive per day that failed to return a positive result for Pss_SMR. All bees that were sampled as day 0 controls were also sampled for pollen counts.

As this was a post-hoc experiment using stock solutions for a different experiment, not all desired samples could be located; no samples of marked bees from day 1 of the second experiment were found. However, as the aim was not to assess the difference in pollen grains between marked and unmarked bees, but rather to assess whether there was a correlation between number of pollen grains and Pss_SMR counts, this sample limitation did not affect the overall results. Each bee sampled for pollen grains also had a known Pss_SMR cfu/bee count from the results of section 2.3.1.

To count pollen grains, 80-µl of the original stock solution from each bee was taken and mixed with 20-µl of Alexander stain. Five counts of kiwifruit pollen from each sample were made using a haemocytometer. All nine squares of the haemocytometer were counted unless the first square had more than 100 pollen grains (in which case only 3 squares were counted) or more than 400 grains (in which case only 1 square was counted). Using the volume of the haemocytometer and the mean number of pollen grains from the five counts, an average number of kiwifruit pollen grains per mL (equivalent to per bee) was obtained for each sampled bee.


To investigate the correlation between the number of pollen grains and the Pss_SMR cfu count per bee, both datasets were log transformed, following a correction to eliminate zero-counts from the datasets. For the pollen counts, 0.5 was added to the average number of pollen grains per count for each individual bee before the average was then converted into a per ml figure, representing a theoretical limit of detection. For Pss_SMR cfu counts, zero counts were substituted with the limit of detection (33 cfu/mL).

For each of the two experiments, the correlation for each day was assessed for significance, and then a general linear model (GLM) was used to assess whether there was a significant difference in slope between day zero and the other sample days. The two experiments were then directly compared to each other with all days combined using the similar GLM.


3 Results

3.1 Temperature Comparison

Temperature and RH were recorded for 13 days in beehives in the field and in the containment facility, and for seven days in ambient laboratory conditions (Figure 1). Temperature in the brood of hives was maintained at or just below 35°C both in the field and in the containment facility. Temperature fluctuated daily in the three other locations, with considerable overlap in the temperature ranges between them.

Elevated temperatures in the containment facility meant that the outer edge of hives experienced higher temperatures than would be expected in the field. The temperature record from the outer edge of the hive in the field had 13 out of 633 (2.0%) data points above 30°C, with a total time above 30°C of 6.5 h. The hive in the containment facility had 49 of 617 (7.9%) data points above 30°C, with a total time above 30°C of 24.5 h.

Figure 1. Temperature records in the brood and outer edge of beehives in the field and containment facility, and ambient laboratory environment over the course of the experiments.

There was no significant difference between the relative humidity of brood in the field (mean = 45.8% RH) and in the containment facility (mean = 46.1% RH, P = 0.3604, two-tailed t-test, Bonferonni correction; Figure 2), or between RH in the outer edges of hives in the field (mean = 66.0% RH) and in containment (mean = 64.8% RH, P = 0.0306, two-tailed t-test, Bonferonni correction; Figure 2). Mean RH was significantly different between all other comparisons (P<0.0001, two-tailed t-tests, Bonferonni correction; Figure 2), but the RH in ambient laboratory conditions (mean = 50.9% RH) was midway between brood and outer recordings for both field and containment beehives (Figure 2).


Figure 2. Relative humidity in the centre of brood (‘Brood’) and the outer edge (‘Outer’) of beehives in the field (‘Field’) and in the PC 1+ containment facility (‘Tent’), and in ambient laboratory conditions (‘Lab_Ambient’). Boxes show 25th and 75th percentiles and medians, error bars show 5th and 95th percentiles and dots show outliers.

3.2 Selection of Psa analogue

In laboratory tests, both Pss_SMR and Psa survived for more than nine days when inoculated onto plastic discs, although Psa declined 10,000-fold (R2 = 0.97, P<0.0001, one-tailed t-test of significance of regression coefficient of log-transformed data) over the nine days, while Pss_SMR declined just 10-fold (R2 = 0.23, P=0.0231, one-tailed t-test of significance of regression coefficient of log-transformed data). Because of time and resource constraints, the experiments proceeded with this strain of Pss_SMR despite its higher survival on plastic discs at room temperature, as we considered that the testing of a worst-case scenario for bacteria survival was preferable to selecting a weak Pss strain that might not survive as well as Psa.

Figure 3. Mean number of colony forming units (CFU) of streptomycin-resistant Pseudomonas syringae pv. syringae (‘Pss’, red squares and red best-fit line) and Pseudomonas syringae pv. actinidiae (‘Psa’, blue diamonds and blue best-fit line) on plastic discs in ambient laboratory conditions.


3.3 Survival and transmission of Pss_SMR in beehives

3.3.1 Pss_SMR survival and transmission by pollen foraging bees

In the initial experiment that involved feeding contaminated pollen to bees, 176 bees were marked over approximately 6 h. Counts of marked bees in hives after 24 h found 54 bees in Hive No. 2, 15 in Hive No. 1, and just one marked bee in Hive No. 4. Because of this low number and the unequal distribution of marked bees, we only present data here from Hive No. 2, as well as the initial bacterial counts from day 0.

No Pss_SMR bacteria were found on bees that were caught before the contaminated pollen was presented. Bees that had foraged on the contaminated pollen and collected enough to fill their pollen sacks carried mean bacteria numbers of 1.1 x 107 CFU/bee (Figure 4). When pollen sacks were removed from the bees before testing, the average count fell significantly to 1.6 x 105 CFU/bee (P = 0.014, one-tailed t-test, Bonferonni’s correction for multiple comparisons; Figure 4). Other pollen foragers caught at the entrance of hives during the experiment with different pollen, showed a mean of 1.0 x 103 cfu/bee, which was significantly lower than the amount on bees without pollen sacks that had foraged directly on the contaminated pollen (P = 0.012, one-tailed t-test, Bonferonni’s correction for multiple comparisons; Figure 4), but the range of recorded bacterial counts overlapped for these two groups, indicating that the Pss_SMR had been spread rapidly to other bees that had been foraging elsewhere.

Figure 4. Mean number of colony forming units (CFU) of streptomycin-resistant Pseudomonas syringae pv. syringae recovered on day 0 from bees that had been foraging on contaminated pollen (with and without pollen sacks), and from bees that had been foraging on a different pollen source. Boxes show 25th and 75th percentiles and medians, error bars show 5th and 95th percentiles and dots show outliers.

The mean number of CFUs of Pss_SMR on marked bees in Hive No. 2 after 24 h was 10,000 times lower than those recorded on bees whose pollen sacks had been removed after foraging on contaminated pollen on day 0 (Figure 5). Pss_SMR was still detected on one marked bee from this hive on Day 13.


Figure 5. Mean number of colony forming units (CFU) of streptomycin-resistant Pseudomonas syringae pv. syringae bacteria on marked bees from Hive No. 2 following presentation of contaminated pollen to pollen-foraging bees. Day 0 values are from bees that had their pollen sacks removed before testing for bacteria.

The second experiment to assess the transmission of bacteria among bees within a hive successfully coated 100 marked bees from each of five hives with pollen contaminated with high numbers of Pss_SMR. Both Hive No. 5 and Hive No. 4 showed evidence of streptomycin-resistant bacteria on one of the control bees sampled for background numbers of bacteria (Figure 6); the maximum number recorded was 58 CFU/bee. Day 0 bees coated with contaminated pollen had a mean of 3.2 x106 CFU/bee. The mean CFU/bee count had dropped 100 fold to 6.3 x 103 after 24 hours (Figure 6). Only two hives had marked bees with Pss_SMR by day 9, with the highest count from one bee of 7.2 x 102 CFU/bee.

Unmarked bees were found to be carrying Pss_SMR within 24 h of the inoculation of marked bees with contaminated pollen (Figure 6). The presence of Pss_SMR on unmarked bees peaked at 3.1 x 103 CFU/bee on Day 2, and one hive had two bees with Pss_SMR on day 5. Pss_SMR was not found on unmarked bees on day 9.


Figure 6. Mean number of colony forming units (CFU) of streptomycin-resistant Pseudomonas syringae pv. syringae (Pss_SMR) on marked bees coated with Pss_SMR-contaminated pollen (‘treated’) and unmarked control bees (‘control’) in the five hives. Lines show means for each treatment per sampling day

3.3.2 Pss_SMR survival in beehives

The mean bacteria count from bees inoculated on day 0 was 1.4 x 105 CFU/bee. On day 9, no Pss_SMR was found on bees from any of the five hives, and only Hive No. 1 on day 6 had a mean population of 21 CFU/bee from the outer edge of the hive (Figure 7). In the centre of the brood, Pss_SMR was last found on day 2 in Hive No. 1 at a count of 6 CFU/bee (Figure 7).

The plastic discs inoculated with Pss_SMR had a mean of 4.1 x 107 CFU/disc on day 0. Plastic discs from the outer edges of hives had a mean population of 2.4 x 105 CFU/disc remaining on day nine, with a maximum number recorded of 5.3 x 105 CFU/disc (Figure 7). In contrast, no Pss_SMR was recorded on day 9 from discs located in the centre of the brood, and on day 6 the mean population was 9.3 x 103 CFU/disc (Figure 7).

Pss_SMR was not detected on control (not inoculated) discs on day 0, but was found on control discs in hives on days 1, 2 and 6. The peak population recorded was 1.1 x 104 CFU/disc from Hive No. 1 on day 2.


Figure 7. Mean number of colony forming units (CFU) of streptomycin-resistant Pseudomonas syringae pv. syringae in the outer edge and centre of brood of beehives (Nos 1-5) in the field on plastic discs and on caged bees.

3.4 Survival of Psa in beehives

A wild-type strain of Psa was used to inoculate caged bees and plastic discs in two hives that were kept inside a tent in a containment facility. Inoculated bees showed an initial population on day 0 of 1 x 105 CFU/bee from one sample of five bees, but no Psa was found on the other two day 0 samples. This is probably because the drops of inoculum had not completely dried before the bees were frozen in -80°C, which could have caused the droplets to fall off the bees. This methodological problem was avoided in the Pss_SMR trial, and would not have been an issue on any of the following sampling days.

After 24 h, Psa was not found on any of the discs or bees that were located in the centre of the brood (Figure 8). Psa was last found on discs in the outer edge of Hive No. 1 on day 2 at a count of 4.3 x 103 CFU/disc, and on outer bees the same hive on day six at a count of 6 CFU/bee (Figure 8). No Psa was found on discs or on bees on day 9 or day 14.


Figure 8. Mean number of colony forming units (CFU) of Pseudomonas syringae pv. actinidiae in beehives (Nos 1 and 2) in containment on plastic discs and on caged bees.

3.5 Survival of Psa on kiwifruit pollen at 35°C

Following results from section 3.4 showing that Psa could not be found on discs or on bees in centre of the brood after 24hrs, but could be found in the outer edges of the hives, we conducted an experiment to compare survival of Psa at 35°C in an incubator to Psa kept in laboratory conditions. Due to the presence of Pss_SMR at days 9 and 13 on bees that had either been covered in or had collected contaminated pollen (respectively), we also compared survival of Psa in these two temperature conditions on kiwifruit pollen.

Figure 9. Mean number of colony forming units (cfu, solid line) of Pseudomonas syringae pv. actinidae on plastic discs in; (A) ambient laboratory conditions and, (B) 35°C incubator. Diamonds represent individual counts from discs, and the dashed line is the limit of detection.

A B


In this experiment, Psa did not survive on discs at ambient temperature for as long as was recorded in section 3.2 (Figure 9A). Psa was last found on discs kept at ambient temperature on day 4 (Figure 9A), and on disc kept at 35°C on day 2 (Figure 9B). Due to the small sample sizes it is not possible to assess the significance of the difference in survival times between the two temperature conditions, but these data do show that Psa can survive at 35°C. This means that the rapid loss of Psa from discs kept in the centre of brood in beehives (section 3.4) cannot be attributed to the effects of temperature alone.

The same isolate of Psa was used to contaminate three samples of kiwifruit pollen. While one sample of pollen showed results that were markedly different from the other two, Psa was still detectable in all three samples on day 8 in both the 35°C incubator and in ambient laboratory conditions (Figures 10A and 10D). Although statistical analysis is not possible on these data, the results appear to show that this same isolate of Psa survived better on pollen than it did on the plastic discs, and there is still no support for the hypothesis that the high brood temperature led to the rapid decline in Psa populations within the beehives (section 3.4).

Figure 10. Number of colony forming units (cfu) of Pseudomonas syringae pv. actinidae on three different contaminated pollen samples (red, green and blue lines) in; (A) ambient laboratory conditions and, (B) 35°C incubator. Dashed purple line is the limit of detection.

3.6 Correlation between Pss_SMR survival and pollen-grain retention

When broken into separate days (including separating honey bees with and without pollen sacks into day 0 and day 0.5 respectively), the correlations of pollen grains/bee and Pss_SMR cfu/bee from the first experiment are only significant on day 0.5 (R2=0.75, P<0.001) and day 1 (R2=0.42, P=0.012; Figure 11). The correlations from day 0 (R2=0.78, P=0.113) and day 2 (R2=0.01, P=0.824) were not significant (Figure 11). There was no difference in slope between day 0 and the other days; day 0.5 (SED=1.44, P=0.138); day 1 (SED=1.19, P=0.676) or day 2 (SED=1.23, P=0.186).

For the second experiment, data from day 1 were removed from analysis as we were not able to obtain data from marked bees on that day, although the data are shown in Figure 12. None of the correlations from individual days was significant on its own: day 0 (R2=0.691, P=0.166); day 2 (R2=0.001, P=0.847); or day 5 (R2=0.21, P=0.37; Figure 12). The slope of day 0 was not significantly different from the slopes of day 2 (SED=1.56, P=0.164) or day 5 (SED=1.6, P=0.294).

A B


Figure 11. Correlations between the number of pollen grains per bee and the number of colony forming units (CFU) of Pseudomonas syringae pv. actinidae (Psa) at days 0 (blue), 0.5(purple), 1 (red) and 2 (green) after pollen-foraging honey bees were presented with kiwifruit contaminated with Psa.

Figure 12. Correlations between the number of pollen grains per bee and the number of colony forming units (CFU) of Pseudomonas syringae pv. actinidae (Psa) at days 0 (blue), 1(red), 2 (green) and 5 (purple) after honey bees were coated in Psa-contaminated kiwifruit pollen.


Because we could not find a significant difference between slopes within the two experiments, we proceeded to compare the overall slopes between the two experiments. The correlations from both experiment 1 (R2=0.86, P<0.001; Figure 13) and experiment 2 (R2=0.77, P<0.001; Figure 13) were significant, but there was no significant difference in slope between the two experiments (SED=0.187, P=0.476).

Figure 13. Correlations between the number of pollen grains per bee and the number of colony forming units (CFU) of Pseudomonas syringae pv. actinidae (Psa) from pollen-foraging honey bees that were presented with kiwifruit contaminated with Psa (red line and squares, ‘Experiment 1’), and honey bees coated in Psa-contaminated kiwifruit pollen (blue line and diamonds, ‘Experiment 2’).


4 Discussion

This project was successful in isolating a streptomycin-resistant strain of Pss that could be used as a proxy for Psa for field experiments. Although Pss_SMR survived better in the laboratory on plastic discs than Psa, this was preferable to using weaker doubly resistant strains of Pss that would have led to underestimates of survival times of Pseudomonas bacteria in hives. For future experiments of this type, we recommend developing multiple strains of bacteria to be used as proxies so that the most appropriate strain can be used, but in this project there was not enough time to conduct such an experiment.

By the ninth day of the experiments, neither Psa nor Pss_SMR could be found on bees that were inoculated with a droplet of bacteria, suggesting that the bacteria had died or had been cleaned off at some point between six and nine days after inoculation. In contrast, Pss_SMR continued to survive on plastic discs in the outer edge of hives at day 14, while discs kept in the centre of the brood showed no sign of Pss_SMR on day 9. Some of the bees that had been coated with pollen contaminated with Pss_SMR still had the bacteria on day 9 and day 14.

These results show that plant pathogenic bacteria such as Pss and Psa are able to survive in beehives and on bees for periods up to two weeks (Pss) and potentially longer, although the population drops sharply over the first day or two. The risk of bees spreading Psa will depend on the initial inoculation loads in infected orchards, and the amount of inoculum needed to start an infection on a vine as a result of deposition by honey bees on flowers.

While the hives kept in tents in the containment facility showed slightly elevated temperatures in the outer edge of the hives compared with hives in the field, our results show that the brood temperature of 35°C is unlikely on its own to be responsible for the rapid decline in Psa populations in that experiment. There is evidence to suggest, however, that Psa and Pss might survive better on pollen than on bees or discs, and therefore the persistence of Pss on bees that had been coated in contaminated pollen might be explained by the retention of small numbers of kiwifruit pollen grains. These data suggest that if Psa is found to contaminate pollen in orchards, the survival of Psa in beehives is likely to be closely tied to the retention of kiwifruit pollen grains on pollen-foraging bees.

There is a need to further investigate what factors other than temperature might contribute to the decrease in the survival of Psa in beehives, and there is also a need to assess the risk of infection from honey bees carrying inoculum loads of 10 – 1000 CFU/bee.

The rapid spread of bacteria through the hive to surfaces and to bees that had not been directly contaminated shows that bees are potentially capable of spreading the disease among orchards without the hive being moved.

4.1 In conclusion

1. Bees that have been contaminated with Psa through visiting infected flowers could contaminate other bees from their hive that are foraging in another orchard, possibly several km away. In the worst case scenario, it is theoretically possible for a single hive to spread PSA bacteria over 10 km without being moved.


2. Bees moved between orchards overnight could still be carrying viable Psa when they visit flowers the following day.

3. The number of Psa bacteria on bees inside a hive will quickly decline after the hive is removed from the contaminated orchard, although it is possible for bees still to be carrying very small numbers of bacteria for 14 days afterwards.

It needs to be noted, however, that it is not known whether the numbers of bacteria carried by bees are capable of causing infection in an orchard.

5 Recommendations

5.1 Honey Bee Management Plan

There are several risk pathways that need to be managed:

5.1.1 The risk of bees from stationary hives moving Psa between orchards

There is no known method of preventing this occurring. The risk of this type of spread means that there is no rationale for preventing the movement of hives within an individual Priority Zone (PZ).

5.1.2 The risks of bees spreading Psa as a result of moving hives between orchards

The discovery of Psa in early spring 2011 at sites outside of the initial infected zones suggests that Psa may already be more widely distributed than thought. Thus there is a case to be made for treating all hives as being ‘potentially contaminated’ with Psa. Consequently, we recommend that the practice of moving hives between orchards should be discouraged in regions that are currently Psa-free and also between separate high-risk areas. We also recommend that hive movement into an orchard in a currently Psa-free region from an infected orchard, PZ or high-risk area should be actively prevented.

These movement controls could be achieved by:

1. Providing all kiwifruit growers with pollination contract templates that require that the colonies have not been used for kiwifruit pollination in a previous orchard in the same season

2. Holding meetings for growers and beekeepers to explain the potential risks of bee vectoring of Psa. A list of all beekeepers and growers attending should be kept and the beekeeper list compared with ZESPRI’s list of beekeepers providing pollination hives, so that they can be contacted individually.

3. Carrying out auditing of hive movements by tracking individual hives and beekeeper operations.


5.1.3 The risk of hives being moved from infected orchards to within flight distance of another orchard

No colony from an infected orchard or PZ should be moved to within 5 km of an orchard outside the infected zone, by:

1. Providing growers with the location of all apiary sites so they can observe hive movements near their orchard.

2. Restrictions on hive movements could be mandated by MAF for biosecurity reasons. We recommend requesting this from MAF.

3. Providing beekeepers with the location of all orchards and a map marked with the 5 km radius. Ask beekeepers to identify their apiaries that currently lie within a 5-km buffer zone now, so that they can plan for what to do when hives are moved out of kiwifruit orchards.

4. Encouraging growers to keep all flowers in the sward in their orchards mown, to reduce the chance of other bees foraging in their orchard.

5. Holding grower and beekeeper meetings as discussed above, and contacting growers and beekeepers who do not attend.

5.2 Possible Legislative Tools

If the powers of the Biosecurity Act (1993) were to be used, it would be possible to have all hives used for pollination numbered and only allow them to be moved to pre-approved apiary sites until after the end of flowering. This could then be audited for compliance. However, this is likely to require the payment of compensation as stipulated under the Biosecurity Act (1993).

6 Acknowledgements

The authors wish to thank Joel Vanneste for his contribution to the design of this study, Harold Henderson for his assistance in data analysis, and Deirdre Cornish and Warren Yorston for their technical assistance.

7 References

Walker R, Rossall S, Asher MJC 2004. Comparison of application methods to prolong the survival of potential biocontrol bacteria on stored sugar-beet seed. Journal of Applied Microbiology 97: 293-305.

Documents

Survival of Pseudomonas syringae pv. actinidiae P. s. pv. syringae