Efficacy of Bumble Bee Disseminated Biological Control Agents for Control of Botrytis Blossom Blight of Rabbiteye Blueberry

Abstract

Botrytis blossom blight caused by Botrytis cinerea may cause severe crop loss in rabbiteye blueberry, necessitating applications of expensive fungicides. Commercial bumble bees, Bombus impatiens, were tested as vectors of the fungicidal biological control agents, Prestop® (Gliocladium catenulatum) and Mycostop® (Streptomyces griseoviridis), against blueberry blossom blight. A single bumble bee hive and four flowering blueberry plants were confined within each of ten 1.5-m³ insect exclusion cages. Test products were applied onto the bodies of worker bees exiting each hive through a tubular V-shaped dispenser containing the test product affixed to the hive's entrance. Dissected floral parts were plated onto agar and recovery frequency of biological control agents was determined. Stylar infection rates were 100% for S. griseoviridis and 70% for G. catenulatum. In field trials, bumble bees vectored a UV-fluorescent dye and biological control agents to about one-third of the open blueberry flower clusters in a small field in as few as 8 days. Flowers were inoculated with B. cinerea 10 days after placing biological control agent vectoring bee hives in the field. The inoculated flowers on bee-visited stems (Prestop treated flowers) had more white corollas (71%) and a lower disease incidence than unvisited flowers (52% white corollas). These results indicate that bumble bees can vector sufficient biological control agent to blueberry flowers to reduce floral damage caused by B. cinerea, however, disease control may not have sufficient economic impact under high disease pressure.

Keywords:

• Bombus impatiens
• Vaccinium ashei
• Botrytis cinerea
• Prestop
• Gliocladium catenulatum
• Mycostop
• Streptomyces griseoviridis

INTRODUCTION

Blueberries are a major North American fruit crop, and their production has steadily increased, particularly in the southeastern United States, where the area planted in rabbiteye blueberries (Vaccinium ashei Reade) more than doubled between 1982 and 1992 (Moore, 1993). One of the major causes of crop loss of rabbiteye blueberries in the Gulf South in recent years has been death or injury of flowers due to late spring freezes, blossom diseases, or a combination of both. Botrytis blossom blight (caused by the fungus Botrytis cinerea Pers.: Fr.) may cause severe crop loss of rabbiteye blueberries but usually is unimportant on highbush blueberry (Bristow and Milholland, 1995; Eck, 1989; Milholland and Meyer, 1995; Smith, 1998; Varney and Stretch, 1966). The fungus attacks blossoms, tender green twigs, and leaves in early spring, causing symptoms that are often mistaken for other blossom diseases or for freeze injury. Infected flowers and twigs quickly turn brown or black and then die. The fungus produces abundant gray masses of conidia that can quickly spread throughout the field. High humidity (>95%) and cool temperatures (15 to 20°C) are ideal for Botrytis infection. Often the fungus will advance from infected flower clusters into the stem, girdling it and killing all flowers above the infection point. Botrytis is an aggressive saprophyte which can quickly invade tissue injured by a spring frost (Austin, 1994). The same fungus also causes Botrytis fruit rot, but in the southeastern U.S., losses on rabbiteye blueberry are usually greater due to Botrytis blossom blight than Botrytis fruit rot.

Botrytis blossom blight has been especially severe in Georgia on the rabbiteye cultivar, Tifblue, when cool, rainy weather occurred during bloom (Krewer et al., 1986). Due to the similarity between Botrytis blossom blight and freeze injury (Austin, 1994; Bristow and Milholland, 1995; Krewer et al., 1986; Milholland and Meyer, 1995; Varney and Stretch, 1966), some losses attributed to freeze injury in the past may have actually been caused by Botrytis blossom blight. Fungicides recommended for control of Botrytis blossom blight include cyprodinil + fludioxonil (Switch 62.5WG), fenhexamid (Elevate 50WDG), captan + fenhexamid (CaptEvate 68WDG), pyraclostrobin + boscalid (Pristine WG), ziram (Ziram 76DF), and captan (Captan 50WP) (Krewer et al., 2010); however, fungicidal control of Botrytis blossom blight often is unsatisfactory. This could be due to the application of the wrong fungicide because the disease was misidentified or because the fungicides were applied at the wrong time. Fungicides have to be applied at the proper stage in flower development for control of Botrytis blossom blight to be successful. ‘Tifblue’ flowers are susceptible to Botrytis blossom blight when they are at or near full bloom (stages 5 and 6) (Smith, 1998). Weather conditions during bloom may prevent the application of fungicides at the correct time. In addition, fungicides are expensive to purchase and to apply.

The use of biological agents (BCAs) for disease control is an alternative to conventional chemical fungicides. BCAs avoid many of the risks associated with conventional fungicides including damage to the environment and exposure of workers to fungicides and the public to fungicide residue on crops (Fravel, 2005), and the chances of pathogens developing resistance to BCAs is also lessen (Sutton et al., 1997). BCAs lend themselves to delivery by insects such as bees. Honey bees and bumble bees have been used to deliver bacterial and fungal biocontrol agents to flowers (Johnson et al., 1993; Maccagnani et al., 1999; Peng et al., 1992; Thompson et al., 1992; Yu and Sutton, 1997). Dedej et al. (2004) equipped honey bee hives with dispensers containing the commercial product Serenade (Bacillus subtilis) and achieved significantly less mummy berry disease. Kovach et al. (2000) found that inoculum of Trichoderma harzianum delivered to strawberry flowers by bees provided better disease control than when delivered by spray applications. A number of BCAs have been shown to reduce infections of flowers by B. cinerea, thus, potentially reducing latent Botrytis fruit infections. In trials in New Zealand, isolates of Gliocladium and Clonostachys were shown to be highly effective at reducing B. cinerea conidiophores on strawberry flowers (Card et al., 2004).

Bees are the primary pollinators of blueberry flowers and are also known to be able to vector BCAs from the hives to the flowers. The use of bees as the means of application for BCAs and chemical fungicides could provide better flower protection than spray applications. The same characteristics that make native bees, particularly bumble bees, excellent blueberry pollinators (i.e., buzz pollination, fidelity for Vaccinium, and large hairy bodies) also makes them good candidates for the precise delivery of biological fungicides to blueberry flowers. Bumble bees as alternative pollinators of blueberry crops are increasing in popularity as concern grows over the reduced global availability of strong honey bee colonies. Additionally, honey bees are reluctant foragers of blueberries and have a strong propensity for abandoning the crop (Sampson and Cane, 2000). Female bumble bees, on the other hand, are faithful blueberry pollinators that possess specialized hairs on their hind legs, which are used to carry pollen as larval provisions. In addition to pollen, these hairs can also accommodate a diversity of non-nutritional substances that have the consistency of pollen. Some of these peculiar particles have included flour, sawdust, coal dust, as well as fungicidal dusts and fungal spores (Wahl, 1966; Roulston and Cane, 2000). Bees acquire these inedible substances passively from inside their colonies or during their foraging trips. Thus, analogous particles, which include fungicides formulated as fine dusts, can be dispensed onto those forager bees entering and exiting the hive, and the bees can then vector fungicidal particles precisely to the more vulnerable crop blooms at a lower cost and with negligible waste and environmental contamination (Kovach et al., 2000; Janisiewicz and Korsten, 2002). Bees can effectively deliver fungicides or BCAs at the precise time that the flowers are most susceptible to Botrytis infection (stages 5 and 6). Colonies of bumble bees are used solely as sources of efficient blueberry pollinators and their stores of honey and pollen are not intended for human consumption; therefore, bumble bees are perfect candidates for the safe conveyance of specially formulated fungicides to blooms.

The main objective of this study was to determine if commercially available bumble bees (Bombus impatiens) could be used to carry biological control agents (BCA) and conventional fungicides to blueberry flowers in sufficient quantity to protect the flowers against blossom blight disease caused by Botrytis cinerea.

MATERIALS AND METHODS

Inoculum

Botrytis cinerea isolates were obtained from blueberry by harvesting conidia with a sterile needle from diseased flowers and spreading conidia over the surface of potato dextrose agar (PDA) acidified with 1 ml of lactic acid per liter. The isolate was maintained on PDA until needed. Fungal inoculum was prepared from 10- to 16-day-old cultures grown under fluorescent lights at approximately 22°C. A conidial suspension was prepared by flooding plates with sterilized distilled water and dislodging conidia with a glass rod. Conidia were counted, the spore concentration was adjusted to 2.5 × 10⁵ conidia per ml, and flowers on test plants or stems were inoculated using a hand pump sprayer. Control plants and stems were sprayed with sterile distilled water. Following inoculation plants were incubated for two days in a moisture chamber at 20°C and 100% relative humidity (RH).

Disease Assessment and Data Analysis

Two weeks after inoculation, flowers on each plant or stem were rated for disease development on a visual scale of 0 to 7, where 0 = no visible symptoms, 1 = small red spots on corolla, 2 = one to four water-soaked lesions on corollas of flower cluster, 3 = corollas of all flowers in a cluster brown, ovaries green and undamaged, 4 = corollas brown and ovaries of young developing berries damaged, 5 = all flowers or berries within a cluster dead, 6 = all flowers or berries within a cluster dead and covered with spores, 7 = lesion extends from dead cluster into stem. Data were subjected to analysis of variance by PROC GLM with PC/SAS software (SAS Institute, Cary, NC, USA). Means were separated by least significant difference (P = 0.05).

Cage Study

One small commercial bumble bee (Bombus impatiens) hive (Koppert B.V., the Netherlands) was placed inside each of ten insect cages (2 m sq × 2 m h). Fourteen blueberry plants with open flowers were placed inside each cage. Bumble bees were allowed to exit the hives through Tylon tubes placed in the exit holes. A test product was placed in each exit tube and refilled each morning. Four products were delivered: UV fluorescent dye, the commercial fungicide, Switch® 62.5WG (a mixture containing two active ingredients—cyprodinil, a systemic component and fludioxonil, a contact component; Syngenta Crop Protection, Inc., Greensboro, NC, USA), and two commercial biological control fungicides [Prestop® (Gliocladium catenulatum J1446) and Mycostop® (Streptomyces griseoviridis Strain K61) Helsinki, Finland]. A fifth hive with nothing in the exit tube served as the untreated control. There were two cages per treatment with fourteen plants in each cage. Blueberry plants in the cages with the florescent dye treatment were observed after dark each day using a UV light. The presence of the fluorescing dye on the stigmas of all open blueberry flowers confirmed that the bees carried the product to the flowers. Plants were removed from the cages and inoculated using a hand pump sprayer with conidial suspensions of B. cinerea. Control plants sprayed with water instead of inoculum were included in each set of inoculations. Styles were collected from each bush in the control and two BCA treatments and plated out on acid PDA + Triton (Prestop and control) or nutrient agar (Mycostop) to determine if the biocontrol fungus was present on the style.

Fluorescent Dye Field Studies

Two field studies were conducted to determine if and how far the bees would carry the fluorescent dye. One study was conducted at the research station in Poplarville, Mississippi and the other was conducted at an organic blueberry farm located in Stone County, Mississippi. Two hives of bees were placed on stands about 1 m high in the center of each field. Fluorescent dye was placed in the exit tube of each hive. Flowers were examined after dark in the field using a UV light. The number of bees exiting and returning to the hives at the Poplarville location was observed at 5- to 10-min intervals. Approximately 25 flower clusters were collected from randomly selected branches on 15 bushes located from 10 to 40 m (avg. 23 m) from the hives at each of five dates and examined in the laboratory using a UV light.

Prestop Field Study

A field study was conducted to determine if bees would carry the BCA product, Prestop, to blueberry flowers in sufficient amount to reduce disease. Two bumble bee hives were placed near 30 bushes of the rabbiteye cultivar, Tifblue, growing in a single 25-m row located at a residence in Poplarville. Prestop (∼0.4 g) was placed in the exit tube of each hive each morning. Flowers were collected from six bushes located at various distances (from 2 m to 20 m) from the hives, and styles from these flowers were plated out on acid PDA + Triton to determine if G. catenulatum was present on the styles. Ten days after the first hive was placed in the field, 40 flowering stems were collected from each of 6 blueberry bushes. Twenty stems from each bush were collected from inside an insect exclusion bag which had been placed over part of the bush before the hives were placed in the field. Bees could not visit the flowers within the bagged areas of each bush so these stems were considered the untreated controls. The other 20 stems from each bush were located outside the bags and their flowers were available for the bees to visit so these stems were considered the Prestop treated flowers. Within each set of 20 stems, 12 were randomly assigned to be inoculated with B. cinerea and 8 were assigned to be sprayed with water. Each stem was placed in a 150 mm × 25 mm tissue culture tubes filled with sterile distilled water, inoculated with B. cinerea, or sprayed with water, and incubated at 25°C in a moist chamber at ∼100% RH for 5 days. One day after inoculation, each stem was evaluated for the overall color of the corollas. The total number of brown, semi-brown and white corollas of each of the 240 stems was recorded. Three days after inoculation each cluster was given a disease score.

RESULTS AND DISCUSSION

Cage Study

Observations using UV light confirmed that the bumble bees in the fluorescent dye treatment cage disseminated the dye to the stigma tips of nearly 100% of the flowers at stages 5 and 6. This demonstrated that the bees did obtain the dye on their bodies when they exited the hive through the tube containing the dye and that they deposited the dye on the flowers they visited. There were no significant differences in the amount of disease development among the different treatments, probably because most of the flowers in each cage were severely damaged by the large number of bees in each cage. Streptomyces colonies grew from 100% of the styles collected from the Mycostop treatments, and Gliocladium colonies grew from 70% of the styles collected from the Prestop treatments. No Streptomyces or Gliocladium colonies grew from the control treatment.

Fluorescent Dye Field Studies

Flower stems were collected from bushes at the organic blueberry farm and observed using a UV light to determine the presence of the dye. The percentage of clusters with at least one fluorescing flower increased from 6 to 14 to 17 to 28 to 32% with each of the five collection dates (Table 1). There was not a significant difference due to distance from the hive in the percentage of flowers which fluoresced. The bush with the most pink fluorescing clusters was located about 4 m SSW of the hive with the pink dye and the bush with the most white fluorescing clusters was located 19 m WNW of the hive with the white dye. At the Poplarville location, the percentage of clusters with flowers which fluoresced also increased at each collection date: from 10 to 22 to 29% for an average of 19% (Table 2). There was not a significant difference due to distance from the hive in the percentage of flowers which fluoresced. These results indicate that the bees did carry the dye to the flowers under field conditions. When averaged over ten observation periods on 2 days the average number of bees exiting the hives were 0.3 bees per minute and the average number returning to the hive was also 0.3 bees per minute (Table 3).

Table 1 Bumble Bee Dissemination of Fluorescent Dye at an Organic Blueberry Farm Located in Stone County, Mississippi, USA ^z

Sample date	Average number flowers/stem	Flowers fluorescing (%)
March 21	26.4	5.6
March 23	21.2	13.8
March 25	24.4	17.2
March 27	17.3	27.9
March 29	19.1	31.5
Average	21.7	19.2
^zTwo hives were placed in the field on 19 March 2008 at an average distance of 23 meters from each bush. Two additional hives were placed in the field on 25 March 2008. Fluorescent dye was put in the exit tube of each hive each day.

Table 2 Bumble Bee Dissemination of Fluorescent Dyes from Two Hives (One with Yellow or Pink Dye in Exit Hole; the Other with White Dye in Exit Hole) Located in a Research Blueberry Field at Poplarville, Mississippi, USA (March 2008)

	UV field count		March 21, # of clusters		March 23, # of clusters		March 25, # of clusters
Plant	# Open clusters	# Fluorescing	Collected	Fluorescing	Collected	Fluorescing	Collected	Fluorescing	Total, % fluorescing
1	339	52	33	8	19	14	20	15	51
2	91	0	10	0	23	2	19	0	4
3	173	1	15	0	27	2	18	1	5
4	226	2	26	0	36	3	17	2	6
5	173	1	20	1	29	2	21	7	14
6	152	36	20	0	24	3	18	9	19
7	117	51	14	8	25	6	21	14	47
8	102	16	13	1	27	7	20	4	20
9	172	9	20	0	24	5	21	2	11
10	52	2	11	0	13	2	20	2	9
Total	1,597	170	182	18	263	57	195	56
Average, %		11		10		22		29	19

Table 3 Observations of Bees Exiting and Returning to Hives Containing Fluorescent Dye in the Blueberry Field at Poplarville, Mississippi, USA on March 20 and March 23, 2008

March 20, 2008					March 23, 2008
	Observations		Number of bees			Observations		Number of bees
Hive with	Started	Ended	Exiting	Returning	Hive with	Started	Ended	Exiting	Returning
Yellow dye	16:25	16:30	8	2	Pink dye	12:20	12:30	0	2
	16:45	16:55	4	2		12:40	12:50	2	2
						16:08	16:18	0	0
						16:28	16:38	0	0
White dye	16:35	16:45	1	4	White dye	12:30	12:40	2	2
	16:55	17:05	5	6		12:50	13:00	1	4
Total	35 min		18	14		60 min		5	10

Prestop Field Study

Two, 5, 8, and 10 days after the hives containing Prestop were placed in the field, Gliocladium was recovered from 6, 10, 32, and 12% of the styles, respectively (Table 4). Gliocladium was recovered from the styles of all the sampled bushes, with the greatest number being from the bushes located the furthest distance (10 and 20 m) from the hives. The uncovered flower clusters had significantly fewer brown (10%) and semi-brown (19%) corollas and significantly more white (71%) corollas than the flower clusters from stems that were bagged (23% brown, 24% semi-brown, and 52% white) (Table 5). Six days after inoculation, the uncovered stems had a lower disease score (1.9) than the covered stems (2.3). As expected, inoculated flowers had more brown (26%) and semi-brown (31%) corollas, less white corollas (43%), and higher disease scores 3 and 6 days after inoculation than those flowers that were not inoculated (2.3% brown, 7.7% semi-brown, and 90% white). There were significant interactions between control treatments (+/−) and inoculation treatments (+/−) based on the percentage brown, semi-brown, and white colloras and the disease score 6 days after inoculation. Among the stems inoculated with Botrytis for 6 days, those that were bagged, had significantly more diseased corollas that were brown (37%) or semi-brown (34%) and significantly fewer white healthy corollas (28%). Unbagged blooms on open stems that were accessible to the bumble bees exhibited less disease (15% brown, 28% semi-brown, and 57% white). There were no differences in the corolla color or disease score among the stems inoculated with water. These results indicate that the bees carried Prestop to the flowers that were not bagged and that the Prestop reduced the damage due to Botrytis infection.

Table 4 Recovery of Gliocladium (Prestop) from Styles of ‘Tifblue’ Blueberry Flowers Collected from the Field 2, 5, 8, and 10 Days after First Release of Bumble Bees Carrying Prestop Fungicide ^y

Flower collection date	Days after 1st bee release	Number of clusters collected	Number of styles plated	Styles with Gliocladium (%)
March 21, 2008	2	19.33 ^z	103.83 ^z	5.64
March 24, 2008	5	19.50	108.50	10.30
March 27, 2008	8	22.67	121.17	31.74
March 29, 2008	10	28.67	143.67	12.28
^yStyles were placed on acid potato dextrose agar containing Triton, and the number of styles with Gliocladium colonies was counted 4–6 days after plating.
^zAverage number from six replications.

Table 5 Main Effects and Interactions of Corolla Color (Percentage) and Disease Score Six Days after Inoculation of Detached Blueberry Stems with Botrytis cinerea, and the Significance of Main Effects and Interactions on the Four Experimental Parameters for Treatments Involving Application by Bees of a Biocontrol Agent (Prestop) to Blueberry Flowers (Factor A) and Their Subsequent Inoculation with Botrytis (Factor B) as Determined by Factorial Analysis of Variance

Treatment	N	% Brown corollas		% Semi-brown corollas		% White corollas		Disease score ^z
Bagged (Factor A −)	120	23.2	a ^y	24.4	a	52.4	b	2.3	a
Prestop (Factor A +)	119	10.0	b	18.8	b	71.3	a	1.9	b
LSD		5.3		5.1		5.9		0.3
Inoculum
Botrytis (Factor B −)	143	26.3	a	30.9	a	42.8	b	2.7	a
Water (Factor B +)	96	2.3	b	7.7	b	90.1	a	1.4	b
LSD		5.4		5.2		6.0		0.3
Interactions
Bagged × Botrytis (−/+)	72	37.4	a	34.2	a	28.5	c	3.1	a
Prestop × Botrytis (+/+)	71	15.0	b	27.6	a	57.4	b	2.2	b
Bagged × Water (−/−)	48	1.9	c	9.8	b	88.3	a	1.2	c
Prestop × Water(+/−)	48	2.6	c	5.6	b	91.8	a	1.5	c
LSD		7.7		7.4		8.5
		% Brown corollas		% Semi-brown corollas		% White corollas		Disease score
Source of Variation	Treatments	F value	P > F	F value	P > F	F value	P > F	F value	P > F
Factor A	Prestop (+/−)	28.7	<.0001	23.4	<.0001	6.4	<.000	7.33	.007
Factor B	Botrytis (+/−)	51.6	<.0001	78.9	<.0001	86.9	<.000	73.0	.0001
A × B		18.6	<.0001	5.8	<.0001	1.4	0.24	13.5	.0003
^zDisease development was rated on a visual scale of 0 to 7, where 0 = no visible symptoms to 7 = lesion extends from dead cluster into stem.
^yMean values followed by different letters in the same column within a group are significantly different as determined by Fisher's protected least significant difference (LSD) at α = 0.05.

CONCLUSIONS

Commercial bumble bees can be used as “living-brushes” to efficiently disperse biological fungicides to susceptible flowers. Simple lightweight dispensers can be affixed to hive entrances. These dispensers are weatherproof and bees soon become accustomed to using them as an egress. Some BCA products seem to be more acceptable to bees than others. Bumble bees used in the study were not seen to remove the BCA product from their bodies after exiting through the dispensers. In fact, almost every floral visit led to stigmas becoming inoculated with the beneficial Gliocladium fungus (Prestop), a fungus that reduced Botrytis infection.

The level of disease control achieved in our study was not high enough to predict that Botrytis blossom blight could be controlled using bees to disperse either a BCA or a conventional fungicide during a year in which weather conditions favored severe disease. However, in fields where bees are being used as pollinators, BCAs or fungicides placed at the exits of hives should reduce the level of infection in an average year. This agrees with the conclusion of Dedej et al. (2004) in which they suggested use of a hive-dispersed biocontrol product (Serenade) as a supplement during pollination to reduce the risk of mummy berry disease. They suggested this as a prudent practice to optimize the benefits to pollination of high bee densities by reducing the associated disease-vectoring risk (Dedej et al., 2004).

ACKNOWLEDGMENTS

We would like to thank Melinda Miller-Butler, Chris Werle, Jenna Thrash, and Rosemary Smith for technical assistance and Koppert for supplying the bumble bee colonies used in these studies. Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the US Department of Agriculture and does not imply its approval to the exclusion of other products or vendors that also may be suitable.