Pollination Biology of Jujubes and Longans and the Importance of Insects in the Pollination of Crops in Vietnam

Hanh Duc Pham

A Thesis presented to The University of Guelph

In partial fulfillment of requirements for the degree of Doctor of Philosophy in Environmental Biology

Guelph, Ontario, Canada

ABSTRACT

POLLINATION BIOLOGY OF JUJUBES (ZIZIPHUS MAURITIANA) AND LONGANS (DIMOCARPUS LONGAN) AND THE IMPORTANCE OF INSECTS IN THE POLLINATION OF CROPS IN VIETNAM

Hanh Duc Pham Advisor

University of Guelph, 2012 Professor Gard W. Otis

The floral and pollination biology of jujubes (Ziziphus mauritiana) and longans

(Dimocarpus longan) were studied near Hanoi, Vietnam. Jujube is a protandrous species with three phases of flowering. After a brief asexual phase, anthers dehisce and release pollen in the afternoon of the day of anthesis. Stigmas become most receptive the following day when flowers are actively secreting nectar. Both jujube and longan flowers are visited during the day by insects of many families, particularly honeybees and flies

(syrphids, calliphorids, and muscids). Honeybees, Apis cerana, made up 84% of floral visitors to jujube flowers and 47 – 95% to longan inflorescences.

Bagging experiments revealed that diurnal insect visitors are very important in fruit production of both jujubes and longans. In jujubes, no fruits were set during the first pollination trial early in the flowering period. Fruit set increased to 0.17% midway through flowering and 2.21% for the trial conducted late in the flowering period. Fruit set recorded one week after anthesis suggested that all types of pollination may result in fruits, but 7 weeks after anthesis only open pollination (unbagged flowers) and diurnal pollination treatments yielded fruits. Most fruits (~97%) were estimated to result from

honeybee visits to flowers. Longans are also predominantly pollinated by diurnal insects

(~84%), but with minor contributions from wind pollination (8.4%) and self-pollination

(7.7%). A. cerana was estimated to contribute 67% of longan pollination.

Pollination requirements for 39 Vietnamese crops were reviewed. Most benefit from insect pollination. For 8 crops important in Vietnamese agriculture for which there were sufficient data, crop yields and values were estimated. Honeybee pollination resulted in ~50% of yields of these 8 crops, contributing ~900 millionion USD of their total values. This analysis indicates that the pollination service provided by honeybees is enormous.

ACKNOWLEDGEMENTS

As an international student, I wish to express my thanks to several persons who have provided support and guidance throughout my study at the University of Guelph.

Firstly, I would like to thank my advisor, Dr. Gard W. Otis, for providing me with valuable mentorship, support, encouragement, and guidance, not only in my study but also in my life at Guelph. I also would like to thank other members of my advisory committee, Dr. Cynthia Scott-Dupree who came to Vietnam to assist me in setting up experiments, Dr. Ernesto Guzman, and Dr. Gary Umphrey, for their support and guidance for my study. Many thanks to Dr. Steve Marshall for his assistance with identification of insects at the study site in Vietnam.

I would like to express my gratitude to the Association of Universities and

Colleges of Canada (AUCC) and Canadian International Development Agency

(CIDA) for providing my scholarship. My study would not have been possible without their support through the Tier 2 CIDA-UPCD Project directed by Dr. Gard Otis. The assistance of several other people was invaluable for completion of this thesis research.

Dr. Dinh Quyet Tam and Mrs. Nguyen Thi Hang at the National Apiculture Joint

Stock Company of Vietnam, and Dr. Phung Huu Chinh at the Bee Research and

Development Center facilitated many aspects of my research and granted me leave from my work to pursue PhD studies. Thank you to Joshua Cowan, who volunteered many hours to meet with me for consultations. Many thanks for statistical assistance from Karen Hand and Michelle Edwards, and computer assistance from Bettina

Martin. Special thanks to the families of Mr. Nguyen Van Thanh and Mr. Nguyen

Cong Hoi, and other longan and jujube growers for allowing me access to their orchards for my experiments. I am also very appreciative of the assistance provided by personnel of the Bee Research and Development Center of Vietnam and student assistants from the Hanoi Agricultural University.

Finally, these acknowledgements are far from complete without recognizing the support, encouragement, and patience of my wife, Tuyen. Her support has not been only in words, but also in deeds as she has been on my behalf looking after our children by herself while I have been in Canada. She has had the patience to be together with me from beginning to end of my study in all of its ups and downs over the past four years.

Table of Contents

ACKNOWLEDGEMENTS ...... IV LIST OF FIGURES ...... X LIST OF TABLES ...... XVI CHAPTER 1. GENERAL INTRODUCTION ...... 1

1.1. Pollination ...... 1 1.2. Pollination in Vietnam...... 3 1.3. Longans and jujubes, two important fruit crops in Vietnam that may require insect pollination ...... 11 1.3.1. Longan ...... 11 1.3.2. Jujube ...... 14

1.4. Objectives ...... 15

CHAPTER 2. THE IMPORTANCE OF INSECTS TO CROP POLLINATION IN VIETNAM ...... 17

2.1. Overview of crop pollination in the world ...... 17 2.1.1. Introduction ...... 17 2.1.2. Pollination in agroecosystems in developed countries ...... 20 2.1.3. Agroecosystem pollination in Asia ...... 22 2.1.3.1. Honeybees ...... 22 2.1.3.2. Status of pollination in Asian agroecosystems...... 24

2.2. Overview of crops in Vietnam...... 25 2.2.1. Crops that benefit from insect visitation ...... 25 2.2.2. Crops that do not benefit from bee visitation ...... 61

2.3. Situation of bee pollination of crops in Vietnam ...... 61

CHAPTER 3. THE FLORAL BIOLOGY OF JUJUBE, ZIZIPHUS MAURITIANA ...... 69

Abstract ...... 69 3.1. Introduction ...... 69

3.2. Materials and Methods ...... 75 3.2.1. Study site and cultivars for study ...... 75 3.2.2. Flowering characteristics and quantity of flowers per cluster ...... 76 3.2.3. Anthesis and floral longevity ...... 76 3.2.4. Pollen availability ...... 77 3.2.5. Stigma receptivity ...... 78 3.2.6. Nectar volume and concentration ...... 79 3.2.7. Statistical analyses ...... 81

3.3. Results ...... 81 3.3.1. Flowering characteristics and quantity of flowers per cluster ...... 81 3.3.2. Anthesis and longevity ...... 84 3.3.3. Stigma receptivity ...... 86 3.3.4. Nectar in flowers...... 90

3.4. Discussion ...... 96

CHAPTER 4. POLLINATION AND FRUIT SET OF JUJUBE (ZIZIPHUS MAURITIANA) ...... 104

Abstract ...... 104 4.1. Introduction ...... 105 4.2. Materials and Methods ...... 108 4.2.1. Study site and jujube cultivars ...... 108 4.2.2. Floral visitors ...... 109 4.2.2.1. Determination of the community of flower-visiting insects ...... 109 4.2.2.2. Pollinator determination ...... 110 4.2.3. Pollination treatments ...... 112 4.2.3.1. Fruit set of jujube ...... 112 4.2.3.2. Influence of insect pollination on crop yield and economic efficacy ...... 113 4.2.4. Statistical analysis ...... 114

4.3. Results ...... 115 4.3.1. Diversity, abundance and pollination efficiency of floral visitors ...... 115 4.3.1.1. Diversity of floral visitors ...... 115 4.3.1.2. Floral visitor behaviour ...... 117

vii

4.3.1.3. Pollen loads of visitors...... 124 4.3.1.4. Index of pollinating efficacy ...... 125 4.3.2. Effect of pollination on fruit set ...... 126 4.3.3. Jujube yields ...... 129

4.4. Discussion ...... 129

CHAPTER 5. POLLINATION AND FRUIT SET OF LONGAN (DIMOCARPUS LONGAN) ...... 135

Abstract ...... 135 5.1. Introduction ...... 136 5.2. Materials and Methods ...... 139 5.2.1. Study sites and longan cultivars ...... 139 5.2.2. Characteristics of flowers and flowering of longan ...... 140 5.2.3. Pollination treatments ...... 141 5.2.4. Floral visitors ...... 144 5.2.4.1. Determination of the community of flower-visiting insects ...... 144 5.2.4.2. Estimation of relative pollinating efficacy of main taxa of insect floral visitors ...... 145 5.2.5. Influence of insect pollination on crop yield and economic efficacy ...... 147 5.2.6. Statistical analyses ...... 148

5.3. Results ...... 149 5.3.1. Characteristics of flowers and flowering of longan ...... 149 5.3.2. Effect of insect visitation on fruit set ...... 153 5.3.3. Diversity, abundance and pollination efficacy of floral visitors ...... 159 5.3.3.1. Diversity of floral visitors ...... 159 5.3.3.2. Floral visitor behaviour ...... 162 5.3.3.3. Pollen loads of visitors...... 169 5.3.3.4. Index of pollinating efficacy ...... 170 5.3.4. Economic value of insect pollination ...... 170

5.4. Discussion ...... 173

CHAPTER 6. DISCUSSION AND CONCLUSIONS ...... 180 REFERENCES ...... 187

viii

APPENDIX ...... 250

Appendix 1 Odds ratios and their confidence intervals for comparisons of the interaction between time of day and jujube cultivar for proportions of branches with insect visitors present during scans. (Odds ratio is the ratio of numbers of visitors at one time interval divided by the numbers of visitors at the preceding time interval; P = statistical probability; 95% CI = 95% confidence interval)...... 250 Appendix 2 Odds ratios and their confidence intervals for comparisons of the interaction between time of day and floral branch positions for proportions of branches with insect visitors present during scans. (Odds ratio is the ratio of numbers of visitors at one time interval divided by the numbers of visitors at the preceding time interval; P = statistical probability; 95% CI = 95% confidence interval)...... 251

LIST OF FIGURES

Figure 1.1 Major beekeeping areas in Vietnam...... 7 Figure 2.1 General structure of a complete flower with both male and female structures (Redrawn from McGregor, 1976)...... 18 Figure 2.2 A litchi orchard with many inflorescences extending from branch tips...... 41 Figure 2.3 A longan tree with terminal inflorescences on the tips of branches...... 43 Figure 2.4 A jujube tree growing in an orchard in Vietnam...... 46 Figure 2.5 Rambutan fruits...... 49 Figure 2.6 A coffee branch with flowers...... 50 Figure 2.7 A dragon fruit plant (Hylocereus undatus) with many wilted flowers...... 53 Figure 2.8 A mango tree in full bloom...... 55 Figure 2.9 A peanut plant in bloom...... 57 Figure 2.10 Cashew tree with flowers and immature fruits...... 59 Figure 2.11 An orange tree in Ha Tinh Province with many mature fruits...... 60 Figure 3.1 A jujube tree. The tree has erect branches near the top and drooping branches near the bottom, with a myriad of tiny flowers in clusters at nodes along branches...... 71 Figure 3.2 Jujube rootstocks, each with several side trunks that have grown from grafted scions. After collecting fruits, the trees are pruned. Intercropping with other crops (e.g., corn and squash) is typical...... 73 Figure 3.3 Inflorescences of jujube (F12 cultivar). Flowers in various stages of development are evident within each cluster. The distance between each pair of adjacent clusters was 1.5 – 3.0cm...... 82 Figure 3.4 Numbers of flowers that opened daily within a flower cluster, for three jujube cultivars, each in two different orchards (n = 48 clusters for each cultivar). Data were standardized so that the first flower in every cluster opened on Day 1...... 84 Figure 3.5 Phases of floral development in jujube. Upper left: Asexual period: A flower starts to open; Upper right: the flower is partly opened; petals are upright; and stamens are covered by petals and anthers have not dehisced; Lower left: Male phase: anthesis has occurred; anthers have dehisced and pollen has been presented. Lower right: Female phase: Stamens drooped onto petals; stigma developed and receptive...... 88 Figure 3.6 (a) Shiny receptive stigma with two lobes in F8 cultivar; (b) brown stigma of “Dao Muon” cultivar flower that is no longer receptive. Note that the petals have shrivelled up and the stamens have drooped downward...... 89 Figure 3.7 Test for stigma receptivity in jujube. Two flowers with several bubbles on top of the stigma

within the drop of H2O2 that indicated the stigma was receptive...... 89 Figure 3.8 Proportion of incipient fruit sets from flowers exposed to pollinators on the three days of flowering. Fruit set data were recorded one week after flowers opened. Day 1: floral branches were exposed to pollinators only on the first day of flowering; Day 2: floral branches were exposed only during the second day of flowering; Day 3: floral branches were exposed only on the third day of flowering; Open: floral branches were exposed to pollinators throughout the entire period the flowers were open (n = 96). Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test)...... 90 Figure 3.9 Nectar was withdrawn from a flower with a micropipette...... 91 Figure 3.10 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 20 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 12:00h, 15:00h, and 18:00h were 33.5°C and 41.5%, 33.5°C and 40%, and 29.0°C and 71% respectively (n = 18 for each time point sample)...... 92 Figure 3.11 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 27 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 75%, 29.0°C and 59%, 33.5°C and 40%, 30.0°C and 53%, and 27.0°C and 71%, respectively (n = 18 for each time point sample). .. 93 Figure 3.12 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 29 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 75%, 30.0°C and 52%, 32.0°C and 38%, 30.5°C and 46%, and 26.0°C and 72.5%, respectively (n = 18 for each time point sample). 94 Figure 3.13 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 1 October, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 76%, 26.5°C and 67%, 30.0°C and 48%, 30.5°C and 51%, and 28.0°C and 70%, respectively (n = 18 for each time point sample). .. 95 Figure 3.14 Mean amount of nectar sugar (mg) per flower at different times of day (n = 24 for each bar) on the second day of anthesis. Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test)...... 96 Figure 3.15 Mean (± SE) of nectar volume (a) and sugar concentration (b) per flower in jujube at different times of day (n=72 for each bar). Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test)...... 102 Figure 4.1 A branch of jujube tree with fruits...... 106 Figure 4.2 Pollination treatments applied to jujube flowers. Upper left: open pollination with flowers continuously exposed to wind and insects; upper right: cheese cloth bag to allow for wind pollination; lower left: mosquito netting bag applied at night or during the day to quantify diurnal

or nocturnal pollination, respectively; lower right: paper bag (PBS International, Non-standard) to quantify self-pollination...... 113 Figure 4.3 The proportion of floral branches with Apis cerana or other insect visitors present at different times of day during September, 2010 (n = 1152 floral branches). The proportion of branches with honeybees present differed between all successive time intervals except for those beginning at 9:00h and 11:00h.The proportion of branches receiving visits by other insect species did not differ temporally (P > 0.05 for all pair wise comparison, Tukey’s tests). Data points are the proportion of floral branches with visitors present, with 95 percent confidence intervals, the same letters indicate no significant difference...... 119 Figure 4.4 The proportion of floral branches with Apis cerana or other visitors present on jujube floral branches in three different positions on trees. The presence of A. cerana differed with respect to each floral branch position. Visits by non-honeybees did not differ between high outer and low outer branches (p = 0.954). Data points are the estimates of proportion of branches on which visitors were observed during scan sampling, with 95 percent confidence intervals (n = 1152 scans of flowering branches in each branch position category). The same letters indicate no significant difference...... 120 Figure 4.5 The proportion of floral branches of the 3 jujube cultivars (Dao Muon, F8 and F12) receiving visits by all species of insects at different times of day. Data points are the proportions of branches on which visitors were observed during scan sampling, with 95% confidence intervals (n = 384 scans of flowering branches in each cultivar/observation period). Within an observation period, the same letters indicate no significant difference...... 121 Figure 4.6 The proportion of floral branches at 3 different positions (high outer, low outer and low inner floral branches) receiving visits by all species of insects at different times of day. Data points are the proportions of branches on which visitors were observed during scan sampling, with 95% confidence intervals (n = 384 scans of flowering branches in each branch position/observation period). Within an observation period, the same letters indicate no significant difference...... 122 Figure 4.7 Mean (± SE) duration of floral visit by honeybees (Apis cerana) and members of three fly families during each floral visit, divided into four temporal observation periods. The 4 main insect taxa that visited jujube flowers differed in the duration of their visits (P < 0.0001). For A. cerana and flower flies (Syrphidae), the duration of each floral visit did not vary significantly throughout the day. The visits by honeybees were consistently shorter than for any of the other taxa of flower visitors. Means with the same letters are not significantly different (P > 0.05, Tukey’s test). Numbers located before letters are numbers of insects studied...... 123 Figure 4.8 Insects visiting jujube flowers. A blow fly (Calliphoridae) foraging for nectar (upper, left) and lapping its foreleg (upper, right). A honeybee, Apis cerana, standing on two adjacent flowers

xii

simultaneously (lower, left) and in contact with the anthers of a flower (lower, right)...... 124 Figure 5.1 A longan tree with numerous erect thyrses extending outward beyond the foliage...... 137 Figure 5.2 Morphology of female floral buds one day prior to anthesis. The floral buds with style partially extruded (indicated with an arrow) would open the next day...... 143 Figure 5.3 Longan floral types. (A) The M1 flower with undeveloped pistil; (B) The M2 flower has an easily seen pistil that is smaller than the pistil of the F flower; (C) The F flower has fully developed pistil with two lobed stigma. Stamens in F flowers were ~1-2 mm in length and always shorter than the pistil...... 150 Figure 5.4 A). Diagram of longan thyrse showing lateral spikes and floral buds. The letters show primary (a), secondary (b), and tertiary (c) axes. A secondary axis is equivalent to a spike. B). Detail of tertiary axis, or dichasium with numbers showing the sequence of floral anthesis. The relative sizes of the buds reflect their approximate sizes on the day of anthesis of the primary flower (redrawn from Nakata et al., 2005)...... 151 Figure 5.5 Primary flowers opening in female phase (left) and male phase (right)...... 152 Figure 5.6 Mean number of developing fruits recorded over time in Gia Lam, 2010. All treatments yielded some fruits, except self-pollinated flowers (the last fruit from the self-pollination treatment was aborted between 29 April and 14 May). By 8, 15, and 29 April, open pollination and open + hand pollination treatments had more fruits than wind pollination and self-pollination treatments. Fruit set from open and open + hand pollination treatments was continuously greater than that from wind pollination treatment to the last assessment 13 July (all unadjusted P values < 0.05). Data are the mean numbers of developing fruits per thyrse, with 95% confidence intervals (n = 9 for wind pollination treatment, n = 11 for open, open and hand, and self-pollination treatments at each time of observation). Within an observation date, the same letters indicate no significant difference...... 154 Figure 5.7 Mean number of developing fruits recorded on different dates from the “Ha Tay” and the “Hung Yen” cultivars in Gia Lam, 2010. Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 18 for the “Ha Tay” cultivar and n = 24 for the “Hung Yen” cultivar). Lower and upper confidence intervals for the “Ha Tay” cultivar on 13 June and 13 July were (0.002431, 45.4253) and (0.000283, 278.16), respectively and lower and upper confidence intervals for the “Hung Yen” cultivar on 13 June and 13 July were (0.001349, 25.0298) and (0.000157, 153.67), respectively. Within an observation date, the same letters indicate no significant difference...... 155 Figure 5.8 Developing fruits recorded at different times in Quoc Oai in 2010. All treatments yielded some fruits at each time of observation. Fruit set from open pollination and diurnal pollination treatments differed with that from self-pollination from the third assessment on 28 April through

xiii

to the last assessment on 12 July (P < 0.05). Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 9 for each treatment at each time of observation). Within an observation date, the same letters indicate no significant difference...... 156 Figure 5.9 Developing fruits recorded at different times in Quoc Oai in 2011. All treatments yielded some fruits up to week 12. Incipient fruits were aborted mostly in the first three weeks. Diurnal pollination and/or open pollination had more fruits than one or more of the three other pollination treatments (nocturnal pollination, wind pollination, and self-pollination) during at least one of the assessments. Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 31 for nocturnal pollination and wind pollination treatments; n = 32 for open pollination and self- pollination treatments at each time of observation; in diurnal pollination treatment: n = 32 for the first four weeks and n = 31 for week 8 (18 June) and week 12 (21 July). Within an observation week, the same letters indicate no significant difference...... 158 Figure 5.10 Developing fruits recorded at different time periods for the “Meo” and the “Tron” cultivars, 2011. Data are the mean numbers of developing fruits, with 95% confidence intervals (in “Meo” cultivar, n = 13 trees; in “Tron” cultivar, n = 19 trees). Within an observation week, the same letters indicate no significant difference. The last two checks were performed on 18 June and 21 July...... 159 Figure 5.11 Mean numbers of Apis cerana and other insect visitors per thyrse during 2010 (a) and 2011 (b) flowering seasons. The pattern of diurnal activity and abundance of honeybees in the two seasons differed. Visits by honeybees peaked in the afternoon in 2010 (a) compared to the morning in 2011 (b). Numbers of other visitors observed on longan flowers in 2011 were much fewer than that in 2010. Data points are the mean numbers of visitors with 95% confidence intervals observed during 286, 191, and 191 scans in each time interval in 2010 and 480 scans in each time interval in the 2011 season. The same letters indicate no significant difference...... 163 Figure 5.12 Mean numbers of insect visitors to thyrses receiving pollination treatments in 2010 observed during scan sampling, with 95% confidence intervals (n = 166 scans of flowering thyrses in each of the open and open + hand pollination treatments). The same letters indicate no significant difference...... 164 Figure 5.13 Number of Apis cerana and other insect visitors recorded on longan thyrses differing in location on the trees in 2011. Floral visitors differed in their visitation to upper and lower thyrses on the trees. Higher thyrses were visited by A. cerana more often than lower thyrses. Other species of insects visited lower thyrses in greater numbers than the higher ones (P < 0.0001). Data points are the mean numbers of visitors observed during scan sampling, with 95% confidence intervals (n = 720 scans of flowering thyrses in each position of thyrse for two groups of insect visitors). The same letters indicate no significant difference...... 165

xiv

Figure 5.14 Mean number of insect visitors per longan thyrse of the “Meo” and the “Tron” cultivars at different times of day in 2011 observed during scan sampling, with 95% confidence intervals (n = 240 scans of flowering thyrses for each cultivar at each time). The same letters indicate no significant difference. Number of visits between each pair of time intervals and within each cultivar did not differ (all P values > 0.05), except for the period beginning at 7:00h (P < 0.0001). Within each time interval and between the two cultivars, number of visits did not differ (all P values > 0.05)...... 166 Figure 5.15 Mean (± SE) duration (sec) of floral visits of four taxa of insect visitors to longan flowers at different times of day. Means with the same letters are not significantly different (P > 0.05, Tukey’s test). Duration of visits for A. cerana, flower flies (Syrphidae), and blow flies (Calliphoridae) did not vary during the day. Houseflies remained on flowers for significantly longer periods of time than the three other taxa...... 167 Figure 5.16 Floral visitors of longan: Top left: A honeybee, Apis cerana; top right: a flower fly; middle left: a blow fly; middle right: a house fly; bottom left: a cockroach; and bottom right: a moth at rest on a floral branch with proboscis extended...... 168

LIST OF TABLES

Table 2.1 Important cultivated crops in Vietnam that benefit from insect pollination, and their economic values...... 27 Table 2.2 Vegetables and fruit crops grown in Vietnam that benefit from insect pollination...... 36 Table 2.3 Important Vietnamese crops that do not benefit from insect visitation in their cultivation. ... 62 Table 2.4 Estimation of the contribution of honeybees on some major crops produced in Vietnam ...... 66 Table 3.1 Mean duration of jujube flowering within flower clusters and quantity of flowers in a cluster (n = 4 trees per orchard) in orchards near Hanoi, Vietnam, 2010...... 83 Table 3.2 Timing of initiation of flower opening of the 3 jujube cultivars grown near Hanoi, Vietnam, 2010*...... 85 Table 3.3 Phases of flowering in the three cultivars of jujube...... 87 Table 3.4 Phases of flowering in jujube...... 100 Table 4.1 Insect visitors recorded on jujube flowers...... 116 Table 4.2 Odds ratios and their 95% confidence limits for comparisons of the proportions of flowering branches of jujube with insect visitors between one time interval and the preceding time interval, during September, 2010...... 118 Table 4.3 Percentage of insects that carried pollen and quantity of jujube pollen on specimens sampled on jujube flowers. (Values include corbicular loads of pollen carried by Apis cerana)...... 125 Table 4.4 Pollination efficacy of honeybees (Apis cerana) and other insect pollinators visiting jujube flowers...... 126 Table 4.5 Effect of bagging experiments on fruit set. Data are numbers of developing fruits recorded, followed by the percentage of flowers studied that set fruits. There were no fruits remaining on trees by week 7 of Replication 1. Noctural and diurnal treatments include effects of wind and self- pollination)...... 128 Table 4.6 Jujube yield and its economic value in Vietnam in 2011...... 134 Table 5.1 Insect visitors captured on longan flowers...... 160 Table 5.2 Percentage of insects that carried pollen and quantity of longan pollen on specimens sampled on longan flowers...... 169 Table 5.3 Pollinating efficacy of Apis cerana and other flower visitors on longan flowers...... 171 Table 5.4 Longan yield in Quoc Oai 2011 and its economic value...... 172

xvi

CHAPTER 1. GENERAL INTRODUCTION

1.1. Pollination

Seeds usually develop when plant ovules are fertilized, which occurs after pollen grains are transferred from the male anthers of a flower to the female stigma of the same or a different flower. This process of pollen transfer is known as pollination. Pollen can be transferred by a variety of animals including bees, other insects, birds, and bats; by wind; and by movements of floral structures that bring pollen into contact with stigmas within the same flower (Faegri and Van Der Pijl 1979).

When pollen from an anther is transferred to the stigma of the same or another flower on the same plant, it is referred to as self-pollination. When pollen from a plant's anther is transferred to the stigma of a flower on a different plant, it is called cross- pollination (Crane and Walker 1984). In some species, not only is cross-pollination necessary, but the pollen must move between two different compatible varieties in order for pollination to result in fertilization and production of fruits (McGregor 1976).

Understanding pollination processes in crop plants is important if crop yields are to be maximized.

Pollination precedes fertilization in all plant species. Through the pollination process, viable seeds are formed and fruit initiated (McGregor 1976). For many crops, pollination increases the numbers, shape, and/or quality of fruits, and therefore is important in helping humans to meet their increasing food demands. Under the forces created by increasing human populations, modern agriculture increasingly relies on intensive land cultivation. As a result of agricultural intensification, natural habitats are

converted to agricultural uses and, consequently, wild bee communities and their contribution to pollination services at the landscape scale are significantly diminished

(Klein et al. 2007). Pollinator diversity and visit frequency to flowers decrease when the distance between the flowers and natural habitats that harbour pollinators increases

(Ricketts 2004; Blanche et al. 2006b; Ricketts et al. 2008). In addition, pesticides also may limit natural pollinator populations by reducing their diversity and numbers (Sahu

2011).

Many insects species are known to provide pollination services, but honeybees

(Apis mellifera L. and to a lesser extent Apis cerana F.) are arguably the most important because they are managed generalist pollinators that pollinate many crop species and wild plants throughout the world (Le Thi et al. 2008) and also because they can be transported to fields and thousands of them can be kept in each hive. The value of honeybee pollination for US agriculture was first estimated by Robinson et al. (1989) and later updated by Morse and Calderone (2000). By estimating the average increase in yields obtained through pollination and the approximate proportion of pollination performed by honeybees, and coupling that with the yields and value of various crops, they estimated that honeybees contributed pollination services worth $14.6 billion USD in 2000. Using a similar method, the economic value of insect pollination for 44 Chinese crops was estimated to be $52.2 billion USD in 2008, or ~25% of the value of those crops

(JianDong and WenFeng 2011). Gallai et al. (2008) showed that pollinators contributed up to $200 billion USD (€153 billion) annually to the yields of crops used by humans for food worldwide, a value that represents ~9.5% of the total value of human food that is produced.

Numerous studies have demonstrated the value of honeybees in pollinating individual crops. For example, in Brazil, the contribution of honeybees to the pollination of soybeans (Glycine max) was studied by Chiari et al. (2005). On a large field of soy beans, three treatments were applied: (1) soybeans open to pollinators (control), (2) soybeans in a 24m2 tent containing a small colony of honeybees; and, (3) soybeans in a

24m2 tent without bees. The results showed that seed yields from plants within the control plot and tent containing bees had yields approximately 58% and 51% greater, respectively, than from the soybean plants within the tent lacking bees. In that study they stated that honeybees accounted for 88% of the pollination achieved by all insects visiting the soybeans. In Chile, a study was conducted on canola (Brassica napus) to evaluate the role of A. mellifera as a pollinator (Araneda et al. 2010). Beehives were placed near a large field at a density of 6.5 hives per ha. Three treatments were applied: open pollination, total exclusion of insects by placing cages (2 x 1 x 2m) with 1mm mesh netting over canola plants, and partial exclusion which involved the use of cages with

2.5mm mesh netting to exclude honeybees and other insects that are the same size or larger than honeybees). The yields from open pollinated plants, total exclusion plants, and partial exclusion were 5.2, 3.5, and 4.7 tonnes per ha, respectively. The yields from the open pollinated plants that were visited heavily by honeybees were ~10.6% higher than from those partially caged to exclude honeybees, translating into a significant increase in yield of 0.5 tonnes of seeds per ha.

1.2. Pollination in Vietnam

Vietnam extends 1650 km from south to north, stretching from 8.45°N to 23.4°N latitude. The climate correspondingly ranges from tropical in the south to subtropical in

the north. This varied climate presents conditions suitable for growth and development of a highly diverse flora, both native and cultivated. Of the commercial crops grown in

Vietnam, many benefit from insect pollination. Such crops include coffee, longan, rambutan, litchi, jujube, cashew, sesame, orange, melon, and cucumber (Phung and Vu

1999; William 2002; Minh 2004; Tran 2004).

The population of Vietnam is ~85.2 million and is growing by more than 1.2 percent per year (General Statistics Office of Vietnam 2008). Both domestic and international demands for food are increasing. In addition, living conditions of the

Vietnamese people have improved remarkably in the past two decades, along with the growing national economy. People tend to demand higher quality food products than in the past. Pollination may be important for production of particular crops as it has often been shown to improve the shape, sugar content, numbers of seeds per fruit, and fruit ripening (Howpage et al. 2001; Mizuno et al. 2004; Gamito and Malerbo-Souza 2006;

Faria-Junior et al. 2008; Kuvanci et al. 2010.)

The only managed pollinators in Vietnam are the honeybees Apis cerana and A. mellifera. A. cerana is endemic to Vietnam and most of Asia. Under natural conditions, colonies of A. cerana nest in cavities in trees, rocks, electrical poles, and brick walls

(Phung and Vu 1999; Oldroyd and Wongsiri 2006). When resources are abundant, bees can cover 6-8 vertical parallel combs (nest volume up to 20L) in spring (~10,000 bees); colony size is much lower when resources are in short supply (i.e., in winter and dearth periods). The species is widely distributed in Vietnam, predominantly in mountain areas, forests, and the buffer zones of natural parks and natural reserves. The Vietnamese call A. cerana “Ong ruoi” (fly-bee) because of its small body size. It is a common visitor to

many commercial crops in Vietnam (personal observation).

Beekeeping with A. cerana started in Vietnam over a thousand years ago (Phung and Vu 1999). Bees were traditionally kept in log hives, although there is ancient evidence that simple top bar hives were occasionally in use. Today, many beekeepers work with A. cerana colonies in movable comb hives that allow for better management and much higher honey yields (Phung and Vu 1999). Traditional beekeeping with log hives and top bar hives still exists, however, and is suitable for villagers in mountainous and remote areas because management systems are simple and financial investment to keep bees is very low. The number of A. cerana bee-hives in Vietnam in 2011 was estimated at ~400,000; honey production per colony ranges widely from 2 to 20 kg per year, depending on the type of beekeeping, skill of the beekeeper, and floral resources near the hives (Dinh, T.Q., pers. comm., 20121).

European races of A. mellifera were first introduced into South Vietnam in the

1960s (Dinh and Pham 2001). They proved to be well adapted to local natural conditions and it was soon evident that this exotic bee species had the ability to produce large honey yields. With the excellent conditions in many parts of Vietnam for development of their colonies and the strong international market for honey exports, beekeeping with A. mellifera in Vietnam developed rapidly, especially in the southern provinces. It has become the most common honeybee species in many regions of Vietnam, with the number of beehives estimated at ~1 million in 2011 (Dinh, T.Q., pers. comm., 20121). A. mellifera also builds several vertical combs in a hive or cavity, with cells, combs and

1 Dinh, T.Q., Vietnam Beekeepers’ Association. Hanoi, Vietnam.

overall colony size all larger than those of A. cerana (Phung 2012). A. mellifera does best where nectar flows are intense. Successful honey production usually involves placing hives near large acreages of crops that secrete copious amounts of nectar (e.g., coffee, longans, rubber). Honey production per beehive is 30 to 40 kg per year. A. mellifera honey accounts for most of the honey produced in Vietnam and contributes nearly 100% of all exported honey (27,000 tonnes in 2011; Dinh, T.Q., pers. comm., 20122).

There are three types of beekeepers in Vietnam, depending on the number of colonies they keep. Professional beekeepers who own and manage between 150 to 3000

A. mellifera hives; they must move their hives several times per year to take advantage of a diversity of nectar and pollen sources. Semi-professional beekeepers with 50 to 150 primarily A. cerana hives that are usually placed around their homes; occasionally the hives are moved to take advantage of crops with good nectar flows or to escape a period of poor resources. Small-scale beekeepers with 2 to 50 A. cerana hives that are kept near their houses.

In total, there are ~30,000 beekeepers throughout the country. Concentrations of beekeeping occur in several southern provinces, the central highlands, the central province of Nghe An, and several northern provinces in the Red River Delta and mountainous northwest (Fig. 1.1).

2 Dinh, T.Q., Vietnam Beekeepers’ Association. Hanoi, Vietnam.

Figure 1.1 Major beekeeping areas in Vietnam.

Beekeeping in Vietnam is facing several challenges. There are several pests and parasites (e.g., wax moths in both bee species; Varroa destructor in colonies of A. mellifera) and diseases (e.g., European foulbrood in both bee species; sacbrood in colonies of A. cerana) (Phung and Vu 1999). Of greater concern to beekeepers is that they are generally not welcomed to place hives in or near particular crops such as longans, coffee, and rice. This is particularly true in Son La Province and the central highlands. Many farmers, when they see many honeybees foraging on the crops, believe that honeybees damage their crops by taking all the pollen to their nests or making the flowers fall off the tree, thereby lowering crop yields. In the case of longan, beekeepers are not allowed to place their beehives near many of the orchards in the Mekong River

Delta because local farmers believe that honeybees transmit longan diseases (Nguyen, N.

T. T., pers. comm., 20123). As with fire blight disease (Erwinia amylovora) of apples, which is spread by honeybees (Ghini et al. 2002), this may be true, but research also indicates that longans are heavily dependent on insect pollinators – specifically honeybees - to obtain good fruit yields (see Chapter 5).

Beekeepers have recently experienced a new and serious problem: residues of carbenzazim, a broad-spectrum fungicide applied to flowering crops to prevent fungal diseases, was detected in Vietnamese honey bound for export (Phipps 2012). Due to extremely low residue limits imposed by the USA, several thousand tons of honey is sitting in ports awaiting export. Failure to sell that honey could force many commercial beekeepers into bankruptcy.

3 Nguyen, N. T. T., Can Tho University, Can Tho, Vietnam.

There are several reasons for studying the pollination of crops in Vietnam. There has never been an analysis of the importance of pollination to crop production in

Vietnam, particularly the pollination that can be attributed to honeybees. Data from other countries is useful, but because of different crop varieties and growing conditions in

Vietnam, that information is not sufficient for a full understanding of the Vietnamese situation. If honeybees prove to be important for crop production, then it is critical to enhance the awareness of farmers and extension workers on the value of honeybees and other insects for pollination of their crops. In Vietnam, there are no commercial pollination services similar to those that can be found throughout developed countries such as Canada and the USA. Crops that require floral visits by insects must rely on natural pollinators plus the free pollination services provided by commercial beekeepers who move their hives to crops for honey production. Not only are they not paid for the service their bees provide, some beekeepers must pay local people for access to floral resources even though nectar and pollen have no inherent value to humans (Do, B.V., pers. comm., 20124). In general, farmers have no concept what would happen to their crop yields if honeybees were not present.

It is difficult to demonstrate to farmers the importance of honeybees as pollinators for their crops because there are many vectors that move pollen between flowers (Faegri and Van Der Pijl 1979), including wild bees and flies, wasps, numerous other insects, and bats. Some crops are pollinated by wind (e.g., rice, corn, and sugar cane (Crane & Walker

1984)), are self-pollinated (e.g., some citrus, cotton, and modern varieties of peanuts)

4 Do, B.V., Thai Binh, Vietnam (+84 984 270 602).

(McGregor 1976; Blanche et al. 2006a; Sykes 2008), or require no pollination at all (e.g., cassava, bananas, and taro root). With farmers growing a diversity of crops in Vietnam that depend on bee pollination to varying degrees, it is no wonder there is confusion.

However, some Vietnamese growers are aware that without pollinators some of their crops set fewer fruits. They apply several methods to facilitate development of fruits, including hand and mechanical pollination, application of chemicals (e.g.,

naphthyloxy acetic acid - a phyto-hormone applied to tomatoes and sweet peppers), or cultivate parthenocarpic cultivars (i.e., cucumbers and squashes) to ensure crop production in the absence of pollinators (Pham, T.K., pers. comm., 20095).

There have been very few studies on pollination in Vietnam. In 2005, the Bee

Research and Development Center of Vietnam (BRDC), in collaboration with the

Southern Fruit Tree Research Institute, assessed the potential adverse effects of transgenic cotton on floral visitors including honeybees (Le Thi et al. 2008). The report from that study listed floral visitors and determined which species should be studied, potential exposure and adverse effect pathways, created risk hypotheses and designed experiments to test those hypotheses.

In 2007, BRDC cooperated with the Hanoi Agricultural Investment and

Development Company to implement applied research to compare effectiveness of pollination effected by honeybees (A. cerana and A. mellifera) and stingless bees

(Trigona sp.) for cucumbers (Cucumis sativus, “Nova” variety) in greenhouses. Yields

5Pham,T.K., Hanoi Agricultural Development and Investment Company, Hanoi,

Vietnam.

from cucumbers when A. mellifera, A. cerana, and stingless bees were present in greenhouses increased ~30%, 24%, and 10%, respectively over the control plots (Pham,

H. D., unpublished).

In 2009 and 2010, BRDC studied the community of flower-visiting insects and determined the effects of honeybee pollination on coffee (Coffea arabica). The results showed that A. mellifera was the dominant floral visitor, making up ~84% of the total visits and augmented coffee yield by ~25% (Truong, T.A., unpublished).

1.3. Longans and jujubes, two important fruit crops in Vietnam that may require

insect pollination

A number of fruit crops in Vietnam benefit from insect pollination (See Chapter

2). Two important fruit crops grown in Vietnam are longans and jujubes, although there is limited information to prove that these crops require insect pollination. There is some evidence to suggest this is the case: (1) research elsewhere suggests that pollination by insects is extremely important for obtaining high yields of fruit; and (2) flowers of these crops are visited by honeybees and other insects in large numbers.

1.3.1. Longan

Longan (Dimocarpus longan; Family Sapindaceae) is an evergreen species that originated in northeastern and southwestern India, Burma, southern China, and Sri Lanka

(Tindall 1994, cited by XuMing et al. 2005; Wong 2000; Paull and Duarte 2011). It is now cultivated widely in tropical Asia, the Philippines, Cuba, Australia, and the United

States (Crane and Walker 1984). In China, longan is reported to have been cultivated for more than 2000 years, with a planting area of 368,200 ha and production of 1.13 million

tonnes in 2007 in the three southern provinces of Guangxi, Guangdong, and Fujian (Gu et al. 2008, cited by Wu 2010).

Thailand is the largest exporter of longan in the world. In 2003, longans accounted for 6.5% of total agricultural production (Vichitrananda and Somsri 2008) with an area planted to longans of 1.6 million ha and annual production of 12.7 million tonnes.

Thailand exported about 1.3 million tonnes, worth more than 39,000 million Bhat (USD

1.27 million). Longan fruits can be eaten fresh or frozen, and can be canned, dried, and processed into drinks, pickles, wine, ice-cream, and yoghurt (XuMing et al. 2005; Lu et al. 2007, cited by Wu 2010).

Longans are grown widely in rural areas of Vietnam and are one of the most important fruits in Vietnamese culture and economy. The total area of commercial plantings of longans is ~98,000 ha (Linh 2008, unpublished), concentrated in the northern provinces of Hung Yen, Son La, Tuyen Quang, and Yen Bai; and the southern province of Tien Giang. The value of longans is estimated at 245 million USD per year. There are several cultivars of longan that differ in the quality of the fruit (Morton and Miami 1987), including Chin Muon Hung Yen (Hung Yen), Chin Muon Ha Tay 1 (Meo), Chin Muon

Ha Tay 2 (Tron), Huong Chi, and Tieu Da Bo (Tran 2004; Tra 2012; Nguyen 2012). The average sizes of the orchards in different areas vary from several hundred m2 to several ha. During flowering, many insects visit the flowers, including bees, flies, beetles, moths, and ants (personal observation). The nectar secreted by longan flowers once harvested by honeybees results in arguably one of the best and highest value honeys in the world on the export market.

In Vietnam, longan is attacked by several pests. The most serious is the stink bug

Tessaratoma papillose (Hemiptera, Pentatomidae). This bug feeds on floral buds and immature fruits, causing them to dry up and drop off. Another important pest is the stem borer, Arbela dea (Lepidoptera, Metarbelidae), which feeds on the bark of the tree near the ground, and then bores into the stem. Trees whose trunks are damaged by this borer fail to grow and affected branches may dry and fall off (Plant Protection Research

Institute 2005).

Longan is also affected by several plant diseases. “Suong Mai” disease is caused by Phytophthora sp. that appears on vegetative buds, flowers, and developing fruits, causing them to stop developing and drop off. The appearance of tiny black spots indicates the presence of this disease. Another disease called “Benh kho hoa nhan”

(drying flower disease) is caused by a fungus, Fusarium sp., which appears during flowering time in late March to April (Plant Protection Research Institute 2005). Floral buds affected by this disease usually dry up and cannot open. This fungus also causes immature fruits to drop in May before harvest.

To control insect pests, longan growers utilize insecticides that include Sherpa®

(cypermethrin; Bayer company) to control T. papillose and Padan® (cartap; Sumitomo

Chemical Vietnam) to control A. dea (Plant Protection Research Institute 2005).

Typically, insecticides are applied before and after flowering to control insect pests.

However, some longan growers apply insecticides when they detect the presence of these insect pests in their orchard, regardless of whether or not the tree is in bloom. This unnecessary application of insecticides affects survival of all beneficial insects, including honeybees, other longan floral visitors and natural enemies of the insect pests.

1.3.2. Jujube

The common name “jujube” refers to several species of fruit trees in the genus

Ziziphus (Family Rhamnaceae). Jujubes are predominantly cultivated in several Asian countries, including India, China, Korea, Thailand, and Vietnam. There are minor acreages in Russia, the Middle East, southern Europe, Australia, and the USA (Reich

1991, cited by William 2002; Hache and Xu 1995, cited by William 2002; Meyer R.

1995; Nguyen 2008; Wanichkul and Noppapun 2009). Jujube trees can tolerate harsh conditions such as drought and high temperatures (Pareek et al. 2007). The fruits are rich in vitamins C, A, and the B complex, and are available fresh, dried, and candied. In addition, the leaves may be fed to cattle and goats (William 2002). Some parts of the plant (flowers, seeds, fruits) can be used for medicinal purposes. The wood is used in the production of musical instruments because it is smooth and strong. Flowers are a source of nectar for honeybees, and the honey is reported to be as good as sunflower honey

(William 2002). In the USA, jujube is sometimes grown as an ornamental plant, but there is also a small commercial jujube fruit industry in California (William 2002).

The use of the common name “jujube” varies between countries. In Vietnam, the jujube species grown commercially is Ziziphus mauritiana which is locally called “Táo gai” (Vu 2000); the same species in India is usually called “Ber”. In contrast, in China the main species grown commercially is Ziziphus jujuba (Liu and Zhao 2009). In China, jujubes have been cultivated for ~4000 years, with current annual fruit production of

~450,000 tonnes (William 2002). In India, jujube is one of the most ancient and common crops, with high economic returns in part due to the low cost of cultivation.

In Vietnam, jujube fruits are highly valued because of their taste and the timing of

harvest. Jujubes are harvested during the time when the Vietnamese celebrate Tet

Holiday (Lunar New Year). They contribute to the diversity of fruits available during this holiday season. Jujubes are grown throughout Vietnam, on riparian land on a commercial scale and in home gardens. The larger plantings (~10,000 ha in total; Dao, X.T., pers. comm., 20126) are in the Red River Delta in the north including districts near Hanoi and the provinces of Hai Duong and Thai Binh), Ninh Thuan province in the south and the

Mekong River Delta in the south - Tien Giang, Dong Thap, and Bac Lieu provinces. Only one crop is harvested per year, with flowering time depending on the cultivar. Jujube nectar is indispensable for A. cerana bees in the Red River Delta during August to

October, a time of year when few other nectar sources are available for honeybees. Many professional and semi-professional beekeepers with A. cerana colonies place their hives near jujube orchards for honey production. Jujube honey is highly valued by the

Vietnamese due to its fragrant odour, light yellow colour, and slow granulation. Jujube also provides pollen for A. cerana foragers, although the amount is insufficient for good development of bee colonies (personal observation).

1.4. Objectives

Longan and jujube are important crops for Vietnamese agriculture and economy.

There is a relatively poor understanding of the value of honeybees and other insects in their production, and of their floral biology and specific pollination requirements. In

Vietnam, detailed information about these topics is almost completely unknown. Some of

6 Dao, T.X., Field Crops Research Institute, Hai Duong, Vietnam.

the information on these two crops is sparse and may be contradictory (see chapters 3 and

5). Thus, further studies are needed to fill these gaps in our understanding of the cultivation of longans and jujubes.

The objectives of this study were to:

1) Survey the crops grown in Vietnam and estimate the value of pollinators,

especially honeybees, in their production. Due to the sparseness of information on

pollination from Vietnam, information obtained in studies conducted in other

tropical countries was used in this analysis;

2) Describe and quantify floral biology of jujube and longans, including: a) floral

morphology; b) sex ratio (in longan); c) the number of flowers per cluster (in

jujube); d) floral anthesis, anther dehiscence, stigma receptivity, and flower

longevity; and e) volume and concentration of nectar in flowers;

3) Quantify the numbers of insects visiting longan and jujube flowers, especially

Apis cerana, Apis mellifera, and other common flower visitors; quantify the

amounts and floral origin of pollen carried by the main insect visitors; and

provide basic descriptions of their behaviour on flowers;

4) Using different bagging treatments, determine the type of pollination for both

longans and jujubes, and quantify the importance of Apis cerana in pollinating

these crops; and,

5) Estimate the proportion of the crop value attributable to honeybees by combining

results of the studies on the type of pollination, floral visitors, and pollen loads.

Through these studies and experiments, the importance of honeybees in Vietnamese agriculture, especially for longan and jujube production, will be clarified.

CHAPTER 2. THE IMPORTANCE OF INSECTS TO CROP POLLINATION IN

VIETNAM

Pollination of flowers by insects is an important process that increases yields of many cultivated crops. In developed countries, there has been a substantial amount of applied research conducted on this subject (Robinson et al. 1989; Heard 1999; Klein et al. 2007). However, for most tropical countries, there has been relatively little research on the pollination requirements of crops and the importance of insect pollination.

Additionally, most growers in tropical countries may be uninformed about the pollination needs of their crops. This chapter presents an overview of information about pollination of cultivated crops in Vietnam, especially those that require bee pollination.

2.1. Overview of crop pollination in the world

2.1.1. Introduction

Pollination is the process by which pollen grains are transferred between the male and female reproductive organs of a flower – i.e., from the anthers (male) to the stigma

(female) of the same or another flower. For most plant species, pollination is required for the fertilization of ovules that develop into seeds. Pollen can be transferred in several ways, for example, by bees, other insects, birds, and bats, or wind (Faegri and Van Der

Pijl 1979).

Flowering plants have several different reproductive structures important in pollination. At the ends of the male stamens are anthers that dehisce to expose pollen grains – which contain the male sex cells. A sticky stigma is located at the distal end of

the female pistil. Seeds form from fertilized ovules (within the ovary) at the base of the pistil (Fig. 2.1, redrawn from McGregor 1976).

Stigma

Pistil Anther Stamen

Petal

Nectar dropletOvary Ovule Nectary Sepal

Figure 2.1 General structure of a complete flower with both male and female structures (Redrawn from McGregor, 1976).

In self-pollination, pollen from an anther is transferred to the stigma of the same or another flower on the same plant. In cross-pollination, pollen from an anther is transferred to the stigma of a flower on a different plant (Crane and Walker 1984).

Moreover, pollen must move between two different varieties in order for pollination to result in fertilization and production of some fruits (e.g., apples; McGregor 1976).

When animals such as bees, flies, butterflies, moths, and birds pollinate plants, they are usually foraging for nectar (e.g., carbohydrate source) secreted by nectaries within the flowers or pollen (e.g., source of proteins and lipids), and pollination is usually an unintentional result of these activities. When feeding, the pollinators accidentally rub against the stamens and pollen is transferred to their bodies. Pollination results when they

move to another flower and pollen is transferred onto a stigma (Missouri Botanical

Garden 2009). If the pollen is compatible with the female stigma, the two sperm cells of the pollen grain move toward the ovules within a “pollen tube”. Once there is fusion of one sperm with the egg nucleus, fertilization occurs, leading to eventual formation of an embryo, and seeds and associated fruits are produced. Adequate pollination generally ensures that a high percentage of ovules are fertilized, resulting in high quality full- bodied fruit (McGregor 1976; Dogterom et al. 2000).

The colours of flowers are diverse and often related to the most important pollinators. For example, bird-pollinated plants usually have colourful flowers and an abundance of nectar, but little or no odour. Fly-pollinated and nocturnally pollinated flowers are frequently white or pale in colour (Faegri and Pijl 1979). Bee-pollinated plants are usually yellow or blue and often have ultraviolet patterns that help to guide the insects into the flower. Flowers of wind-pollinated plants are usually small, with a tiny corolla and long stamens and pistils. They have not evolved to attract animal pollinators, and thus are usually dull coloured, unscented, and lack petals (McGregor 1976; Menzel and Shmida 1993).

Honeybees are important pollinators for many reasons. They are generalist pollinators that live in large colonies. Bees visit many floral species in a day and exhibit

“floral constancy” – with individual foraging trips usually limited to a single species of flowering plant. Hives can be easily moved in and out of crops to meet the specific demands of the flowering crop (Morse and Calderone 2000; Amaya-Márquez 2009).

Honeybees tend to be present in large numbers in most agroecosystems because they are managed by beekeepers.

2.1.2. Pollination in agroecosystems in developed countries

Pollination of many crops in developed countries is mediated by insects.

Pollination by insects improves crop yield in 80% of crop plants (Prescott-Allen and

Prescott-Allen 1990; Ingram et al. 1996). It is estimated that globally 30% of human food results from insect pollination (McGregor 1976). In Canada, ~70% of food crop species need insects for pollination. Some important crops pollinated by the more than 1000 species of pollinating insects in Canada include apples, pears, cucumbers, melons, strawberries, raspberries, cranberries, and blueberries (Anonymous 2008a). Bees, because most species feed pollen to their larvae, are an extremely important group of pollinators.

They have a cosmopolitan distribution, with 20,000 to 30,000 species worldwide

(Michener 2000). Honeybees are the most economically important managed pollinators

(Michener et al. 1994). The value of honeybee pollination has been estimated at 10x more than the value of the honey and beeswax combined (43 million CDN vs. 49.6 million CDN) in Canada (Canadian Association of Professional Apiculturists 1995) and

143x more in the United States when the indirect contribution of honeybee pollination of alfalfa that is consumed by livestock is included in the estimate (Levin 1984). In the

United States alone, honeybees contribute an estimated 15 billion USD towards the value of crops that require or benefit from bee pollination (Morse and Calderone 2000).

Calderone (2012) recently updated these data. Although the cultivated area of crops directly dependent on insects for pollination (apples, almonds, etc.) increased from 1992 to 2009 in the United State, the aggregate crop production was fairly consistent.

However, the value of insect-dependent crops did not increase consistently (Calderone

2012). It decreased from 14.29 billion USD in 1996 to 10.69 billion USD in 2001, and

then increased to 15.12 billion USD in 2009. Values of crops attributed to either honeybee or other insect pollinators had a similar pattern. They declined from 11.20 billion USD and 3.09 billion USD from 1996 to 8.33 and 2.36 billion USD in 2001, and then increased to 11.68 and 3.44 billion USD in 2009, respectively (Calderone 2012). In the United States, migratory beekeeping operations transport more than 2 million colonies around the country annually, following the bloom of crops requiring insect pollination (Morse and Calderone 2000). For pollination of almonds alone, the number of colonies required in California in 2012 was estimated to be nearly 1.5 million, with 1 million originating out of state (Parsons 2012).

Modern agriculture increasingly relies on intensive land cultivation, and movement of colonies by beekeepers is a response to the demands of large-scale agriculture. Colonies are brought to crops for pollination during their flowering, and are then removed so growers can continue overall crop health management. If the colonies remained in situ, they would often fail to thrive because little forage would be available after the crop finishes flowering and there are risks of them being harmed or killed following exposure to pesticides (Martin and McGregor 1973; Crane and Walker 1983;

Stefanidou et al. 2003). The demand for crop pollination is increasing. For example, in the United States, the number of colonies rented for pollination increased from 2,04 million in 1989 to 2,5 million in 1998 (an increase of 22.8%). In the same time interval, there was a 57% increase in the value of the increased yield and quality of crops achieved through pollination by honeybees (9.3 billion USD in 1989 and 14.6 billion USD in

1998). Whereas 20% to 25% of that increase is due to inflation, the rest is a result of the increasing human population with its higher demand for foods pollinated by honeybees

(Morse and Calderone 2000). In the single case of almonds in California, the demand for bee hives for pollination has increased from 802,000 in 1992 to 1,48 million in 2010

(Carman 2012).

2.1.3. Agroecosystem pollination in Asia

2.1.3.1. Honeybees

Two honeybee species are managed for honey production in Asia: the European honeybee, Apis mellifera L, and Asian honeybee, Apis cerana F. In addition, several other honeybees species (Apis dorsata, A. florea, A. andreniformis, Apis nigrocincta, and

A. laboriosa) contribute to honey production and pollination in some regions. Beekeeping with the endemic species A. cerana is an important economic sector throughout Asia, especially in China, India, and Vietnam (Oldroyd and Wongsiri 2006; Phung 2012).

Apis mellifera is the primary species used in beekeeping worldwide. Its colonies generally produce honey yields much larger than those of the Asian honeybee species previously mentioned (Nguyen and Pham 1994). However, management of A. mellifera in Asia requires a substantial financial investment by individual beekeepers. A. mellifera colonies may suffer severe damage from Varroa mites (Varroa destructor) that are native to mainland Asia (Guzman-Novoa et al. 1996). In contrast, A. cerana, produces smaller crops of honey and a much smaller amount of royal jelly than European honeybees, and often abandons hives under adverse conditions (e.g., diseases, attacks by wasps, lack of food, extreme climatic conditions). However, they are unharmed by parasitic mites and require little financial investment by beekeepers at start up, and thus are very suitable for sedentary beekeeping, especially for small-scale farmers (Wongsiri and Chen 1995).

Honey production in Vietnam averages 26 kg and 10 kg per colony per year for A. mellifera and A. cerana, respectively (Dinh, T.Q., pers. comm., 20087).

Stingless bees (Meliponini) are relatively diverse and widely distributed in South and Southeast Asia (Sakagami et al. 1990). Three genera are described from the region, with most species in the genus Trigona. Of the 11 species found in Indochina, 10 are in the genus Trigona; there is just one additional species in the genus Hypotrigona.

Currently three stingless bee species are recognized in Vietnam: Trigona (Lisotrigona) carpenteri, Trigona ventralis, and Trigona laeviceps (Chinh et al. 2005). In comparison with Thailand or Malaya, where 22 and 29 species are listed, respectively (Sakagami et al. 1990), diversity of stingless bees in Vietnam is low. However, additional species recognized as occurring in “Indochina” may be discovered in Vietnam with further field work. In Vietnam, stingless bees are widely distributed but are most commonly found in national parks, nature reserves, and buffer zones that separate protected areas from agricultural areas (personal observation). Some people maintain stingless bee colonies, to harvest their valuable honey, beewax, and propolis. Stingless bees also play an important role in the pollination of ~100 cultivated crops and are efficient pollinators of 9 crops – chayote, coconut, carambola, mango, macadamia nuts, camu-camu, mapati, annatto, and cupuacu (Heard 1999).

Carpenter bees (Xylocopa spp. (Anthophoridae) are found in tropical and subtropical climatic zones (O’Toole and Raw 1991; Ho 1993). They nest in the deep tunnels they chew in dead branches, timbers, rafters, and internodes of bamboo.

7Dinh, T.Q., Vietnam Beekeepers’ Association, Hanoi, Vietnam.

Carpenter bees are visitors and pollinators of several plant species – cotton (Pleasants and

Wendel 2010), passion fruits (Delaplane and Mayer 2000c), melons (Sadeh et al. 2007), and Cordia lutea (McMullen 2012). In the Philippines, three species of carpenter bees

(Xylocopa bombiformis, X. bakeriana, and X. chlorina) have been shown to augment fruit set of passion fruit by 44% in open pollination in comparison with caged plots that excluded pollinators (Rodriguez and Cervancia 1999).

The sweat bees (Hymenoptera, Halictidae) are named for their characteristic of often collecting human perspiration. Sweat bees have a cosmopolitan distribution.

Halictus and Lasioglossum are two of the most important genera of sweat bees (O’Toole and Raw 1991). In Asia, including Vietnam, Lasioglossum (Lasioglossum) subopacum and L. (L.) okinawa have been described (Murao 2011). Most species make their nests in the ground, but some species nest in partially decayed wood. Many halictid species display intermediate degrees of social behaviour, while many others are solitary.

2.1.3.2. Status of pollination in Asian agroecosystems

Pollination services provided by beekeepers in Asia vary by country and crop. In general, however, beekeepers are rarely paid for pollination services. There are few incentives for growers to pay pollination fees to beekeepers because they usually have little or no understanding of the pollination requirements of their crops (see section 2.3).

They generally believe that beekeepers are adequately compensated through the sale of hive products (e.g., honey, pollen).

Many factors influence the effectiveness of insect pollinators in Asia. In parts of

Sichuan Province, China, fruit growers use large amounts of pesticides in their apple orchards. Consequently, beekeepers are hesitant to rent their bees to the growers for apple

pollination. For this reason, apple flowers are poorly pollinated by honeybees and growers have had to resort to hand pollination which is ~8x more expensive than bee pollination (Partap et al. 2001). In southern India, where few insects visit cashew flowers

(Anacardium occidentale), the lack of pollinators was recognized as a major constraint to cashew nut production. To overcome this problem, several management techniques to increase fly populations in cashew plantations have been developed and tested (Reddi

2003).

2.2. Overview of crops in Vietnam

2.2.1. Crops that benefit from insect visitation

In Asia, many crops benefit from bees and other pollinators. The following analysis of the crops grown in Vietnam was conducted in order to estimate their relative dependence on insect pollination. Table 2.1 presents information related to major industrial perennial crops. It summarizes production data from Vietnam as well as taxa of pollinators (worldwide) and the reliance of each crop on insects, particularly honeybees, for pollination. Some crops, like coffee and cashew nuts, contribute extensively to international markets (Reuters 2009; VietNamNet 2009).

Table 2.2 shows similar information for vegetables and additional species of fruit trees that benefit from bee pollination. There are no production data on these crops because most people living in rural areas are subsistence farmers. They normally produce what they need for their daily lives so that they need to buy very little from markets. For instance, many families follow the “VAC” model of farming. VAC is an abbreviation used in Vietnam to describe garden, pond, and livestock production. In their gardens,

they grow different kinds of vegetables and fruits. In ponds, they raise several kinds of fish and shrimp. They usually also rear livestock, including buffalos, cows, pigs, chickens, ducks and occasionally honeybees. With such diverse agricultural systems in small landholdings, it is impossible to estimate the areas of many crops listed in the tables. However, these crops play a very important role in the daily lives of almost all

Vietnamese in rural areas. If the actual areas devoted to these many crops could be determined, the areas and the economic values of the various crops that benefit from insect pollination would far exceed those of plantations for which data are available.

Table 2.1 Important cultivated crops in Vietnam that benefit from insect pollination, and their economic values.

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Anacardium occidentale *Free (1976); Wulfrath and Speck, cited by (Cashew) McGregor (1976); Heard et al. (1990); Bees (Apis cerana, A. Freitas and Paxton (1998); Reddi (2000); mellifera, A. florea, A. dorsata, Bhattacharya (2004); Klein et al. (2007) Centris tarsata, and Bombus **General Statistics Office of Vietnam Great 437 302 667 202 spp.); wasps; flies (Chrysomya (2008) megacephala, Sarcophaga orchidea, Ligyra sp., and

27 Musca sp.); butterflies; beetles;

ants (Formicidae) Arachis hypogaea (Peanut) *Heide (1923), cited by McGregor (1976); Bees (A. cerana, A. mellifera, Hammons (1963), cited by McGregor Halictidae, Megachilidae, Little to (1976); Hammons and Leuck (1966); 255 505 607 307 Ceratina tarsata); wasps modest Rashad et al. (1979); Moffett et al. (1986); (Odynerus priesneri) Blanche et al. (2006a); **General Statistics Office of Vietnam (2008) Citrullus spp. (Melon and *McGregor (1976); Spangler and Moffett water melon) (1979); Chandler and Cocke (1981); Bees: (Apis spp., Xylocopa Essential to Lemasson (1987); Williams (1987); Kuti pubescens, Bombus terrestris, great 3 837 121 102 and Rovelo (1992); Ikeda and Tadauchi Melissodes bimaculata) (1995); Kato and Nogueira-Couto (2002); Sadeh et al. (2007); Klein et al. (2007); **Anonymous (2008b)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Citrus spp. (Orange, mandarin, *Free (1976); McGregor (1976); tangerine, lemon) Anonymous (1981); Singh and Garg (1994); Bees: (A. cerana, A. mellifera; Bhatia et al. (1995); Malerbo-Souza and Trigona spinipes and Little 87 662 121 80 Nogueira-Couto (2002); Malerbo-Souza et Tetragonisca angustula); flies al. (2004); Gamito and Malerbo-Souza (Syrphidae); butterflies; (2006); Gogoi et al. (2007); Klein et al. Neuroptera, beetles (2007); **Nguyen (2008) Citrus spp. (Pomelo) *Free (1976); Malaipan et al. (1992);

Bees (Apis spp., stingless and Chacoff and Aizen (2006, 2007) Modest 40 296 425 126 solitary bees); wasps; flies **Nguyen (2008) (Syrphidae); butterflies; beetles

Cocos nucifera (Coconut) *Free et al. (1975); Free (1976); Pardede et Bees (A. cerana, A. mellifera, al. (1986); Forbes and Cervancia (1993); Trigona spp., Plebeia sp.); Modest 135 1,047 NA NA Heard (1999); Meléndez-Ramírez et al. wasps (Polistes spp.) (2004); Da Conceicao et al. (2004); Klein et al. (2007) **General Statistics Office of Vietnam (2008)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Coffea arabica, C. canephora, *Free (1976); Amaral (1972a, 1972b); Raw C. excelsa (Coffee) and Free (1977); Willmer and Stone (1989); Badilla and Ramírez (1991); Roubik (2002); Bees (A. cerana, A. mellifera, Klein et al. (2003a, 2003b); Marco Júnior T. spinipes, Lepidotrigona spp., Modest to and Coelho (2004); Olschewski et al. Creightonella frontalis, 506 961 1,456 1,400,000 great (2006) Partamona testacea, and Xylocopa spp.) **General Statistics Office of Vietnam

29 (2008)

Cucumis sativus (Cucumber) *Free (1976); McGregor (1976); Bees (Apis spp., Scaptotrigona Anonymous (1981); Robinson et al. (1989); aff. depilis, Nannotrigona Cervancia and Bergonia (1991); Ahmad testaceicornis, Exomolopsis (1992); Canadian Association of Great 0.3 3 243 606 pulchella, Xylocopa chlorine, Professional Apiculturists (1995); Santos X. philippinensis, Megachile (2004), cited by Slaa (2006); Klein et al. atrata); butterflies; beetles (2007) (Diabrotica balteata) **Anonymous (2008c)

Cucurbita spp. (Pumpkin and *Verdieva and Ismailova (1960), cited by

Squash) McGregor (1976); Wolfenbarger (1962), Essential 1 9 91 819 cited by McGregor (1976); Michelbacher et Bees (Apis spp., Xylocopa sp.); al. (1964), cited by McGregor (1976); flies; beetles (Diabrotica Wadlow (1970), cited by McGregor (1976); balteata) Free (1976); Spangler and Moffett (1979);

30 Chandler and Cocke (1981); Williams

(1987); Lemasson (1987); Kuti and Rovelo

(1992); Ikeda and Tadauchi (1995); Kato and Nogueira-Couto (2002); Sadeh et al. (2007); **Anonymous (2007a)

Dimocarpus longan (Longan) *Crane and Walker (1984); Boonithee et al. Bees (A. cerana, A. mellifera, (1991); Wongsiri and Chen (1995); Heard Modest 98 578 425 245 Trigona spp.) (1999); Waite and Hwang (2002); Blanche et al. (2006b); **Nguyen (2008)

Gossypium spp. (Cotton) *Shishikin (1946), cited by McGregor (1976); McGregor and Todd (1955), cited Modest 12 16 546 9 Bees (A. cerana, A. mellifera, by McGregor (1976); Wafa and Ibrahim A. dorsata, A. florea, (1960), cited by McGregor (1976); Megachile spp., Melissodes Skrebtsov (1964), cited by McGregor spp.), Xylocopa varipuncta, (1976); Free (1976); Tanda (1984); Bombus spp.); wasps (Polistes Vaissière et al. (1984); Waller et al. spp.); beetles (Mylabris (1985a,b); Guseinov and Ragim-Zade

oculata) (1985); Berger et al. (1988); Ahmed et al. (1989); Vaissière (1991); El-Sarrag et al. (1993); Delaplane and Mayer (2000b); Rhodes (2002); Viraktamath and Nachappa (2004); Sanchez Junior and Malerbo-Souza (2004); Gulati et al. (2005); Van Deynze et al. (2005); Nachappa and Viraktamath (2005a, b); Ganapathi and Viraktamath (2006); Klein et al. (2007); Hofs et al. (2008); Bozbek et al. (2008); **General Statistics Office of Vietnam (2008)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Hylocereus undatus, *Weiss et al. (1994); Dag and Mizrahi H. polyrhizus, and H. costaricensis (Dragon Fruit) (2005); Valiente-Banuet et al. (2007) Bees (A. mellifera, Bombus Little 21 335 360 121 **Anonymous (2007b), Phuong and Anh terrestris); bats (Leptonycteris curasoae and Choeronycteris (2012) mexicana) Litchi chinensis (Litchi) *Butcher (1956); Das and Choudhury Bees (Apis spp., stingless (1958), cited by McGregor (1976); bees); wasps; flies; ants Modest 87 596 121 72 Anonymous (1981); Ortiz-Sánchez and

32 Cabello-García (1991); YaoChun (1993); Bhatia et al. (1995); Wongsiri and Chen (1995); Stern and Gazit (1996); Heard (1999); Abrol (2006a); **Nguyen (2008) Lycopersicon esculentum *Lemasson (1987); Cribb (1990); Cribb et (Tomato) Bees (A. mellifera), Melipona al. (1993); Cribb (1994); Canadian quadrifasciata, Nannotrigona Association of Professional Apiculturists perilampoides, Bombus (1995); Santos et al. (2004), cited by Slaa hypnorum, B.(Thoracobombus) Little 2 150 303 45 (2006); Cauich et al. (2004), cited by Slaa pascuorum, B. vosnesenskii, B. sonorus, B. terrestris, Amegilla (2006); Sarto et al. (2005), cited by Slaa chlorocyanea, A. (2006); Klein et al. (2007) (Zonamegilla) holmesi, **Anonymous (2006) Xylocopa spp.)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Mangifera indica (Mango) *Free (1976); Anderson et al. (1982); Ortiz- Bees (A. cerana, A. mellifera, Sánchez and Cabello-García (1991); Toit stingless bees, Bombus and Swart (1994); Bhatia et al. (1995); terrestris); wasps; flies Great 77 409 303 124 Wongsiri and Chen (1995); Sasaki et al. (Chrysomya albiceps, Lucilia (1998); Heard (1999); Dag and Gazit sericata, Musca domestica) (2001); Klein et al. (2007); Siqueira et al. (2008); **Nguyen (2008) Nephelium lappaceum *Crane and Walker (1984); Uji (1987); (Rambutan) Wongsiri and Chen (1995); Heard (1999);

33 Bees (A. cerana, A. mellifera, Slaa et al. (2006)

Modest 25 244 121 30 Trigona itama, T. nitidiventris, **Nguyen (2008) T. canifrons, T. irridipennis and T. atripes) Persea americana (Avocado) *Free (1976); Kalman (1979), cited by Bees (A. mellifera, Trigona Crane and Walker (1984); Ish-Am and nigra, Nannotrigona Eisikowitch (1990); Vithanage (1990); perilampoides, Allodape Eardley and Mansell (1993, 1994); Toit (1994); Avilán and Rodríguez (1995); microsticta); wasps (Polistes Great 3 40 182 7 spp.); Malerbo-Souza et al. (2000); Delaplane and Mayer (2000a); Wysoki et al. (2002); Can- flies (Rhyncomya forcipata); Alonzo et al. (2005); Afik et al. (2006, hummingbirds 2007); Klein et al. (2007); **VOV-SGGP (2006)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Prunus domestica (Plum) *Hooper (1936), cited by McGregor (1976); MacDaniels (1942), cited by McGregor Bees (A. cerana, A. mellifera, (1976); Mann and Singh (1983); Langridge A. florea, A. dorsata, Bombus and Goodman (1985); Canadian Association fervidus, B. impatiens, B. Great 2 9 303 3 of Professional Apiculturists (1995); Batra tenarius, Andrena hippotes, (1997); Calzoni and Speranza (1998); Osmia bucephala, Partap et al. (2000); Delaplane and Mayer Lasioglossum coriaceum); flies

34 (2000d); Frève et al. (2001); Sapir et al. (Eristalis dimidiatus) (2007); Klein et al. (2007); **Hoang (2006)

Sesamum indicum (Sesame) *Anonymous (1981); Lee et al. (1988); Panda et al. (1988, 1989); Hyung Rae et al. Bees (A. cerana, A. mellifera, (1995); Baydar and Gürel (1999); Patil and A. dorsata, A. florea, Trigona Viraktamath (2000, 2001); Sachdeva et al. iridipennis, Xylocopa spp., Little to (2003); Sarker (2004); Patil and Ceratina sexmaculata, C. 45 28 1,517 42 modest Viraktamath (2005) smaragdula, Megachile spp., Halictus spp., Catullus spp.; **Nguyen (2008); Oanh (2009) wasps (Vespa spp., Polistes spp., Vespula maculifrons)

Crop Total Crop Scientific Name (Common Area Production Value/ Pollination Value References (* for column 1 and 2; ** for name) of Crop and Insect (ha x (tonnes x tonne Requirements1 (million columns 3, 4, and 5) Pollinators 1000) 1000) (USD) USD) Ziziphus mauritiana (Jujube, *Crane and Walker (1984); Singh (1984); Ber) Devi et al. (1989); Jothi and Tandon (1993); Klein et al. (2007) Bees (A. cerana, A. florea, Trigona spp., Ceratina spp.); **Bạch (2008); (Dao, T.X., pers. comm., wasps (Rhopalidia spatulata); 2012) flies (Chrysomya Modest 10 170 273 46 megacephala); flies (Sarcophaga sp., Musca

35 domestica, Stomorhina discolour, Eristalinus arvorum); ants (Camponotus compressus)

1 Crop Pollination Requirements = dependence on insect pollinators cited from Klein et al. (2007, Fig. 2, p. 1) – “Essential = Pollinators essential for most varieties (production reduction by ≥ 90% comparing experiments with and without animal pollinators); Great = Great production increase/animal pollinators are strongly needed (40 – < 90% reduction); Modest = modest production increase/animal pollinators are clearly beneficial (10 – < 40% reduction); Little = Little production increase/some evidence suggest that animal pollinators are beneficial (> 0 – < 10%)”.

Values of crops in columns 5 and 6 were converted from VND to USD (1 USD = 16,485 VND; Source: Vietcombank 11 Dec 2008). The original values in column 5 were in Vietnamese Dong (VND). Some of the data in columns 3,4, and 5 (the ones that have no citation in column 7) were obtained from government websites while others were from websites from unofficial sources.

Table 2.2 Vegetables and fruit crops grown in Vietnam that benefit from insect pollination. Scientific (Common name) Need for References Insect Pollinators pollination Allium cepa (Onion) Kordakova (1956), cited by McGregor (1976); Bohart et al. Bees (A. cerana, A. mellifera, A. dorsata, A. (1970), cited by McGregor (1976); Wojtowski et al. (1980); Orlova et al. (1981); Haslbachova and Kubisova (1982); florea), stingless bees, Bombus terrestris, B. Essential Wojtowski et al. (1982); Kumar et al. (1989); Lorenzon et al. lapidarius, halictids, Osmia lufa, for seed (1993); Ruszkowski et al. (1993); Canadian Association of Hylaeuscommunis); flies (Episyrphus balteatus, production Eupeodes corolla, Eristalinus aeneus, Musca Professional Apiculturists (1995); Matuszak (1995); Heard domestica, Calliphora vicina) (1999); Prasad et al. (2000); Witter and Blochtein (2003); Yücel and Duman (2005); Sajjad et al. (2008); Saeed et al. (2008). Anethum graveolens (Dill) McGregor (1976); Warakomska et al. (1985); Free (1993) 36 Modest

Bees (Andrenidae, Halictidae); flies Averrhoa carambola (Carambola, starfruit) Phoon (1985); Heard (1999); Klein et al. (2007) Great Bees (A. cerana, T. thoracica) Brassica hirta (Yellow/White Mustard) Anonymous (1981); Jay (1990); Singh and Singh (1992); Alam et Bees (honeybees, solitary bees, Osmia cornifrons, Modest al. (1995); Canadian Association of Professional Apiculturists O. lignaria lignaria) (1995); Abel et al. (2003) Brassica rapa (Turnip), B.napus (Rape, Rapeseed) Nikitina (1950), cited by McGregor (1976); Anonymous (1981); Bees (A. cerana, A. mellifera, A. florea, Andrena D'Albore (1984); Singh et al. (1996, 2000); Abel et al. (2003); ilerda, Osmia cornifrons, O. lignaria lignaria, Great Morandin and Winston (2005); Klein et al. (2007) Halictus spp.); flies (Eristalis tenax, Eristalis spp., Trichometallea pollinosa)

Scientific (Common name) Need for References Insect Pollinators pollination Brassica spp. (Cole crops: cabbage, kohlrabi, Free (1976); McGregor (1976); Anonymous (1981); Prabucki et broccoli, cauliflower) Modest al. (1982); Alam et al. (1987b); Verma and Partap (1994) Bees (A. cerana, wild bees); flies

Cajanus indicus, Cajanus cajan (Pigeon pea) Free (1976); Rathi and Sihag (1993); Sihag and Rathi (1994);

Bees (A. dorsata, Xylocopa fenestrata, Megachile Heard (1999); Sekhar et al. (2001); Klein et al. (2007) Little sp., Megachile lanata, Chalicodoma lanata); wasps (Polistes sp.)

37 Capsicum annuum (sweet pepper) Modest Shipp et al. (1994); Meeuwsen (2000), cited by Slaa (2006); Dag

Bees (A. mellifera, Melipona subnitida, Melipona (pollinators and Kammer (2001); Ercan and Onus (2003); Cruz et al. (2004), favosa, T. carbonaria, Bombus terrestris) essential in cited by Slaa (2006); Cruz et al. (2005); Roldan Serrano and greenhouse) Guerra-Sanz (2006); Dag et al. (2007); Faria Junior et al. (2008).

Carica papaya (Papaya) Allan (1963), cited by McGregor (1976); Free 1976; Ortiz-

Honeybees, hawk moth (Sphingidae: Perichares Little Sánchez and Cabello-García (1991); Westerkamp and philetes philetes) Gottsberger (2000); Klein et al. (2007)

Coriandrum sativum (Coriander) Shelar and Suryanarayana (1981); Baswana (1984); Sihag

Bees (A. cerana, A. mellifera, A. florea, A. dorsata, Great (1988); Heard (1999); Klein et al. (2007) T. iridipennis)

Scientific (Common name) Need for References Insect Pollinators pollination

Daucus carota (Carrot) *Hawthorn et al. (1960); Wojtowski et al. (1980, 1982); Tepedino (1983); Alam et al. (1987a); Kolesnik (1987); Galuszka and Tegrek Bees (A. cerana, A. mellifera, A. florea, Halictus (1989); Kumar and Rao, (1991); Wilson et al. (1991); Ruszkowski splendidulus, Lasioglossum spp., Osmia lufa, et al. (1993); Sinha and Chakrabarti (1993); Vasilevskiĭ et al. Megachile rotundata; Bombus terrestris, Hylaeus Modest (1994); Gosek et al. (1995); Matuszak (1995); Ruszkowski and brevicornis, H. styriacus, H.communis); flies Gosek (1995); Abrol (1997); Schittenhelm et al. (1997); Tepedino (Calliphora spp., Lucilia spp., Eristalis tenax, Musca (1997); Abrol (2006b); Sharma et al. (2008) domestica) **Nguyen T. (2009)

38 Durio zibethinus (Durian) Faegri and Pijl (1979); Morton and Miami (1987); Salakpetch et

Bees (A. dorsata); bats (Eonycterisspelaea); birds al. (1992), cited by Klein et al. (2007); Free (1993); Lim and Great Luders (1998), cited by Klein et al. (2007); Westerkamp and Gottsberger (2000); KienHock et al. (2002); Bumrungsri et al. (2009)

Fragaria spp. (Strawberry) Skowronek et al. (1985); Kakutani et al. (1993); Ikeda and Bees (A. cerana, A. mellifera, Plebeia tobagoensis, Tadauchi (1995); Wilkaniec and Maciejewska (1996); Delaplane Tetragonisca angustula, Trigona minangkabau, and Mayer (2000e); Paydas et al. (2000); Partap (2000); Young Modest Bombus terrestris); flies; butterflies; beetles; thrips Duck et al. (2001); Chen and Hsieh (2001); Asiko (2004), cited by Slaa (2006); Malagodi-Braga and Kleinert (2004), cited by Slaa (2006); Zaitoun et al. (2006); Dimou et al. (2008).

Scientific (Common name) Need for References Insect Pollinators pollination Lactuca sativa (Lettuce) McGregor (1976); Partap and Verma (1993); Goubara and Bees (A. cerana, Lasioglossum villosulum Little Takasaki (2003, 2004) trichopse); flies; butterflies Luffa cylindrical (Vegetable sponge) Ramesh et al. (2007) Modest Honeybees (A. cerana) Prunus armeniaca (Apricots) Canadian Association of Professional Apiculturists (1995); Batra Bees (A. mellifera, Osmia cornifrons, O. lignaria Great (1997); Szábo et al. (2000); Tepedino et al. (2007) propinqua) Psidium guajava (Guava) Ortiz-Sánchez and Cabello-García (1991); Heard (1999); 39 Bees (A. cerana, A. mellifera, A. dorsata, A. florea, Modest Vishweshwaraiah et al. (2002); Rajagopal and Eswarappa (2005);

Trigona sp.) Alves and Freitas (2006); Freitas and Alves (2008) Sechium edule (Chayote) Wille et al. (1983); Heard (1999) Bees (Trigona corvina, T. cupira, T. fulviventris, T. Modest fuscipennis, T. partamona) Solanum melongena (Egg plant, aubergine) Free (1976); Abak et al. (1995); Miyamoto et al. (2006); Hikawa Bees (A. mellifera, Melipona quadrifasciata and Miyanaga (2006); Gemmill-Herren and Ochieng (2008) Modest anthidioides, Bombus terrestris), Xylocopa caffra, Macronomia rufipes); cucumber beetle

Need for pollination = dependence on insect pollinators; many of the terms in the middle column are cited from Klein et al. (2007, Fig. 2, p. 1) – “Essential = Pollinators essential for most varieties (production reduction by ≥ 90% comparing experiments with and without animal pollinators); Great = Great production increase/animal pollinators are strongly needed (40 – < 90% reduction); Modest = modest production increase/animal pollinators are clearly beneficial (10 – < 40% reduction); Little = Little production increase/some evidence suggest that animal pollinators are beneficial (> 0 – < 10%)”.

Agriculture is extremely important in Vietnam with 73% of the population working in this sector and 24.7 million ha or ~75% of the total land area of the country under cultivation (General Statistics Office of Vietnam 2008). Further information and the details of pollination requirements of some of the most economically important crops in Tables 2.1 and 2.2 are discussed below.

Litchi: In Vietnam, as many as 33 cultivars of litchi are grown in home gardens and plantations (Papademetriou and Dent 2002). There are three varieties of litchi fruit based on sweetness: sour, sweet, and sweet + sour known as “Vai nho”. Litchi is grown commercially on an area of ~35,000 ha (Papademetriou and Dent 2002), concentrated primarily in three regions of northern Vietnam: Thanh Ha District (Hai Duong province),

Luc Ngan District (Bac Giang province), and Dong Trieu, Yen Hung, and Hoanh Bo districts (Quang Ninh province). In addition, smaller litchi plantations are scattered over

Hai Phong, Ha Tay, Lang Son, and Bac Ninh provinces. The three major regions of litchi production provide thousands of tonnes of fruit annually for domestic and export markets. Every February and March, ~1000 beekeepers move their hives to litchi plantations during the flowering period (Fig. 2.2). Honeybees are the primary pollinators on both small and large commercial plantings, collecting both pollen and nectar. They always touch the stigmas while visiting female flowers (King et al. 1989, cited by

Davenport and Stern 2005), and in doing so transfer pollen between flowers.

Figure 2.2 A litchi orchard with many inflorescences extending from branch tips.

Litchi is highly self-sterile (i.e., insect pollination is critical for production of fruit) (Pandey and Yadava 1970). Bees and hoverflies (e.g., Episyrphus balteatus) are reported to be the most important pollinators of litchi (McGregor 1976; Dhaliwal et al.

1977). In India, fruit set in bee-pollinated litchi plot was 39% to 62% greater than in a control plot (no bee pollination) and fruit yields with honeybee pollination were twice those of the control (Anonymous 1981). In China, litchi yields in plots pollinated by honeybees were ~2.5 to 3.0 times greater than in orchards without honeybees (YaoChun

1993). Litchi trees in the USA pollinated by bees set an average of 99 fruits per tree, while trees enclosed to prevent bee pollination set only 1 fruit (Butcher 1956). A study in

Israel investigated A. mellifera pollination of two commercial litchi cultivars, “Mauritius” and “Floridian”. Pollination was low during the first male flowering (M1) of cv.

“Mauritius”, reaching an elevated level when the pseudo-hermaphroditic second flowering (M2) started. Pollen density on bees collected from cv. “Mauritius” inflorescences was very low during M1 flowering, but increased during M2 flowering, showing that M1 flowering was not an important source of pollen for litchi fertilization

(Stern and Gazit 1996). In Thailand, A. florea was the primary bee species found visiting litchi cultivars, with a foraging density of 9.00 – 36.50 bees per m2. Apis cerana and A. mellifera foraging densities ranged from 2.83 to 9.25 and 0.08 to 2.17 bees/m2, respectively (Somnuk and Suavansri 2005).

Longan: This evergreen fruit tree is grown for its flavourful fruit throughout tropical Asia, particularly in southern China and northern Thailand (Crane and Walker

1984). Longan, which is grown widely in rural areas in Vietnam, is one of the most important fruits in the Vietnamese culture and economy. The total area cultivated in longan is ~98,000 ha (Nguyen 2008) concentrated in Hung Yen, Son La, Tuyen Quang, and Yen Bai provinces in the north, and Tien Giang province in the south. The value of longan is estimated at 245 million USD per year (Nguyen 2008). Several cultivars –

“Huong Chi”, “Chin Muon Hung Yen”, “Chin Muon Ha Tay”, “Meo” and “Tron” – differ in the quality of the fruit produced (Morton and Miami 1987).

Flowering in northern Vietnam occurs from March to April (Fig. 2.3). Flowers emerge in male, bisexual, or mixed inflorescences. Many insects visit the flowers, including bees, flies, and moths (personal observation), and insect visitation is apparently essential for commercial production (Davenport and Stern 2005). Thousands of beekeepers place A. cerana and A. mellifera hives near longan orchards for the honey, which is of superb quality and highly favoured by Vietnamese.

Crane and Walker (1984) reported that longan may be self-incompatible, requiring cross-pollination. If that is true, insects would play a critical role in longan pollination. Apis mellifera is reported to be the most important longan pollinator in

Thailand (Wongsiri and Chen 1995), where it has been suggested that pollination by honeybees increases longan fruit yield by as much as 30% (Crane and Walker 1984). In a study from Queensland, Australia, both A. mellifera and Trigona spp. influenced initial longan fruit set. Trigona spp. visited longan flowers more often in areas where longan trees were closer to natural forests, primarily because stingless bees more often live in these forests and have a smaller foraging range than A. mellifera (Blanche et al. 2006b).

It is likely that both A. cerana and A. mellifera augment fruit yield of longans in Vietnam

(Table 2.1 and 2.2)

Figure 2.3 A longan tree with terminal inflorescences on the tips of branches.

Melon, Watermelon, Cucumber, Pumpkin, and Squash: These crops, belonging to the Cucurbitaceae family, are grown in home gardens, commercial fields, and greenhouses. They are important foods in the daily lives of Vietnamese. Many gardeners grow these crops in greenhouses to take advantage of climate control and to produce fruit in the off-season. However, they are then faced with the problem of pollination (see below). To overcome that obstacle, some grow parthenocarpic varieties

(in the case of cucumbers) that do not require pollination.

Most cucurbit species are monoecious: they produce separate male and female flowers on the vines. Numerous studies have demonstrated that pollen vectors, in most cases bees, must move pollen from male to female flowers to effect fruit set. Honeybees are important pollinators of various types of melon. For example, cantaloupe melon fields in southern Texas, US, provided with on average 0.8 A. mellifera colonies per ha yielded 11 kg melons per ha. Providing 3.9 colonies/ha increased yields to 18 kg per ha, compared with plots caged to exclude honeybees that produced < 10 kg per ha (Chandler and Cocke 1981). For another melon, Citrullus colocynthis, honeybee pollination resulted in significantly more fruits per plant, higher seed weight, and greater seed viability than hand pollination; > 85% of flowers pollinated by honeybees set fruit, compared with <

20% following hand pollination (Kuti and Rovelo 1992). In Australia, renting honeybee colonies for melon pollination has proven to be financially advantageous because of increased fruit yields (Williams 1987). For watermelons grown in a greenhouse and pollinated by honeybees, fruit set was 41% in var. Klondike and 95% in var. Sugar Baby, while no fruits were set from flowers bagged to prevent pollinator visits (Spangler and

Moffett 1979). Lemasson (1987) found that the presence of A. mellifera increased fruit

set, fruit weight per plant, mean fruit weight, and mean number of melons per plant to

18.5%, 2664 g, 621 g, and 4.3, respectively, compared with 2.5%, 1469 g, 491 g and 2.9, respectively, without honeybees. In Brazil, Kato and Nogueira-Couto (2002) evaluated the importance of insect pollination in melons (Cucumis melo) and concluded that A. mellifera was the most frequent floral visitor and most effective pollinator.

Several other taxa of bees pollinate melons. Bees in the genus Peponapis are specialist pollinators of squashes and melons, and are reported to be the main pollinator of cucurbits worldwide (Cane 2005). A study on watermelons in southern Colorado,

USA, found that commercial seed production was dependent on insects for pollination, and that the native bee, Melissodes bimaculata, appeared to be a more efficient pollinator than A. mellifera (Brewer 1974). In a study comparing the carpenter bee Xylocopa pubescens, a native Israeli species, and A. mellifera as pollinators of greenhouse-grown honeydew melons (Sadeh et al. 2007), pollination by both bees resulted in similar fruit mass and seed numbers. However, fruit set was statistically threefold higher with X. pubescens compared to A. mellifera. This study also showed that X. pubescens can effectively pollinate melons in enclosures.

Jujube: Jujubes are planted extensively in the Red River delta in the northern part of Vietnam. They usually bloom in September and October, and provide both nectar and pollen to pollinators (Fig. 2.4). However, pollen of jujubes is not sufficient for good development of A. cerana colonies (personal observation). Some varieties of jujube are reported to be self-fertilizing, but fruits that develop from self-fertilization are usually smaller and often drop prematurely from the trees (Ackerman 1961). Several species of flies and A. cerana are the most important pollinators of jujubes (Singh 1984; Devi et al.

1989), because jujube plantations are far from forests where most other natural pollinators are found. Without A. cerana, jujubes may not be pollinated sufficiently to attain high yields of fruit (personal observation).

Figure 2.4 A jujube tree growing in an orchard in Vietnam.

Tomato: Tomatoes are planted in both fields and greenhouses. Tomatoes are self- fertilizing and normally self-pollinated (Free 1993). Their tubular anthers release pollen when they are vibrated (Buchmann and Hurley 1978). When fully pollinated, fruits are larger and have more seeds. Because of their ability to buzz-pollinate the flowers, bumblebees are very important insect pollinators of tomatoes, especially in greenhouses

(Delaplane and Mayer 2000f). Ikeda and Tadauchi (1995) found that tomatoes pollinated by B. terrestris were more uniform in shape, and contained more seeds, flesh and vitamin

C than did fruits from plants treated with hormones. Bombus terrestris in Eurasia and B. impatiens in North America have become critically important in pollination of greenhouse tomatoes. Benefits to growers include lower production costs, increased yields, and improved fruit quality (Velthuis and van Doorn 2006). However, in Vietnam,

Bombus spp. inhabit only a small region in the extreme north; consequently, they are not important pollinators of tomatoes in the country as a whole.

Pollination efficiency varies between pollinator species. Apis mellifera may be used to pollinate greenhouse tomatoes, especially during the winter months (Cribb 1994;

Sabara and Winston 2003; Higo et al. 2004). The quantity, and especially the quality, of fruits are increased by honeybee pollination (Cribb 1990; Cribb et al. 1993). However, honeybees do not perform well in greenhouses overall because they fly towards the glass.

Also, honeybees visited relatively few tomato flowers because they could not sonicate the flowers to release the pollen. They visit tomatoes in greenhouses when they have no other floral choice. Furthermore, honeybee efficacy as pollinators of greenhouse tomatoes may vary with growing conditions and varieties (Delaplane and Mayer 2000f). In contrast, bumblebees work well under greenhouse conditions. In England, Banda (1990) compared the effectiveness of honeybees, bumblebees (Bombus terrestris and B. lapidarius), mechanical vibration, and combinations of these in pollinating greenhouse tomatoes; bumblebees achieved higher pollination of greenhouse tomatoes than mechanical vibration, whereas honeybee pollination was little or no better than no pollination. The fruit set and the yield of tomatoes grown in the presence or absence of honeybees did not differ in Lemasson’s (1987) study. For these reasons, greenhouse production of tomatoes focuses on the use of bumblebees as pollinators.

In Australia, carpenter bees (Xylocopa (Lestis) spp.) can be a potential alternative to bumblebees for pollination of tomatoes in greenhouses. When female carpenter bees visited and sonicated flowers in a greenhouse, the tomatoes were heavier and contained more seeds than tomatoes that were not pollinated by Lestis (Hogendoorn et al. 2000).

Currently, there are no managed pollination services for tomatoes in Vietnam. To obtain higher yields in greenhouse tomatoes, gardeners use chemicals (i.e., ß-naphthyloxy acetic acid, a phytohormone) to stimulate development of fruits (Pham, T.K., pers. comm., 20098). Data concerning the area of tomato cultivation is unavailable because most farmers grow tomatoes in small plantings. Further research is necessary under

Vietnamese growing conditions.

Rambutan: Together with litchi and longan, rambutan belongs to the

Sapindaceae family. Rambutans (Fig. 2.5) are produced in the southern provinces of

Vietnam, flowering from November to March, and sometimes into May depending on the cultivar. Honeybees (A. mellifera) collect large quantities of nectar from the flowers, which are either male or bisexual. Lim (1984) determined that fruit formation depends on insect pollination. The major pollinators of rambutan are honeybees (A. cerana and A. mellifera) and stingless bees (Trigona spp.) (Uji 1987; Heard 1999).

8Pham,T.K., Hanoi Agricultural Development and Investment Company, Hanoi,

Vietnam.

Figure 2.5 Rambutan fruits.

(http://www.montosogardens.com/nephelium_lappaceum_r_9_fruit_small.jpg)

Coffee: In Vietnam, three species of coffee are planted in the southern and central highland provinces: Coffea arabica, C. canephora (also called C. robusta), and C. liberica var. dewevrei (Fig. 2.6). Coffee, which provides both nectar and abundant pollen, is one of the most important perennial crops in Vietnam because of its economic value

(Table 2.1). In 2007, the area of coffee cultivation was ~506,000 ha, but that varies from year to year depending on the world market. The total average production of coffee from

2005 to 2010 was 977,000 tonnes (General Statistics Office of Vietnam 2012) and

Vietnam recently emerged as the second largest coffee-producing country in the world

(after Brazil) (Doutriaux et al. 2008).

Figure 2.6 A coffee branch with flowers.

Coffee is pollinated by bees (A. cerana and A. mellifera) and other insects

(Table 2.1). Amaral (1972a) showed that C. arabica pollinated by bees sets ~82% more fruit than coffee not pollinated by bees. Free (1976) experimented on plots of coffee (C. arabica) caged with honeybees, caged without honeybees, or left without caging (i.e., open-pollinated). Coffee plots caged with bees produced 27% to 29% more beans by weight than did open-pollinated plants, and 76% to 107% more beans by weight than did plants caged without honeybees. In a study by Raw and Free

(1977), yields of berries from coffee bushes caged with honeybees were statistically much greater than those from bushes caged without bees, and the yield of mature berries was greater by approximately 100% for plants caged with honeybees. In Costa

Rica, A. mellifera were the most abundant insects on coffee (C. arabica) flowers

(Badilla and Ramírez 1991). Uncaged branches (i.e., available to insect pollinators) in their study produced 16% more berries than caged branches, and berries from uncaged branches were also larger in size and heavier in weight. Similarly, in a study in Brazil, in plants available to pollinators, fruit production was 39% and 168% higher in 1993 and 1994, respectively, than in plants covered to exclude insect pollinators (Malerbo-Souza et al. 2003). Ramalho et al. (1990), Carvalho and Krug

(1996), Malerbo-Souza and Nogueira-Couto (1997) (all cited by Marco Júnior and

Coelho 2004) verified that A. mellifera was the main visitor of coffee flowers (i.e.,

89% of total visits). Chloralictus spp. (Halictidae), Xylocopa spp. (Apidae) and

Tetragonisca angustula (Apidae) were also collected in the experimental plots. Also important in coffee pollination in Brazil are two other bee species, Melipona quadrifasciata (Carvalho and Krug 1996) and Trigona spinipes (Marco Júnior and

Coelho 2004). In Panama, the non–native A. mellifera play the most important role in coffee pollination, but native species may also be involved. Apis mellifera augmented pollination and boosted crop yields by > 50% (Roubik 2002).

In Indonesia, Klein et al. (2003a) studied the importance of insects, wind, and spontaneous self-pollination and degree of self-fertility both in self-sterile Coffea canephora and self-fertile C. arabica. In both species, the highest fruit set resulted from open pollination and cross-pollination by hand. With wind pollination (including self-pollination), fruit set was 16% lower than with open pollination in C. canephora and 12% lower in C. arabica. In self-pollinated flowers and unmanipulated controls, fruit set was extremely low:  10% in the self-sterile species and of 60% and 48%, respectively, in the self-fertile species. Thus fruit set was significantly increased in

coffee cross-pollinated by bees (i.e., honeybees, stingless bees and solitary bees), both in self-sterile and self-fertile coffee species. Willmer and Stone (1989) found that the solitary bee Creightonella frontalis was primarily responsible for pollen transfer in C. canephora in Papua New Guinea, with windborne pollen grains making up only about

10% of the cross-fertilization of the crop.

A diverse and abundant bee community plays a significant role in fruit set.

Numbers and diversity of bees visiting lowland coffee (Coffea canephora) in

Sulawesi, Indonesia, were positively correlated with fruit set of open-pollinated flowers (compared to manually cross-pollinated flowers; Klein et al. 2003b). In the specific agro-forestry systems Klein et al. (2003b) studied, fruit set was higher (95%) when the bee community included 20 or more species compared to just six species

(70% fruit set). Coffee yields in various regions depend on forests providing nesting sites for bees. Integration of agro-forestry with coffee production (i.e., shade-grown coffee) helps to maintain bee diversity. It is estimated that the economic consequences of deforestation, in terms of loss of pollination service to coffee growers in Indonesia and Ecuador, would be ~$47 – 49 USD per ha of forest destroyed (Olschewski et al. 2006).

Dragon fruit: Dragon fruit is a cactus that is widely cultivated in Southeast

Asia, but originated in Latin America (Daniel et al. 2010). It was imported into southern provinces of Vietnam ~50 years ago (Vu 2000) (Fig. 2.9). In late 1989 and early 1990, dragon fruit began to be exported to Taiwan, Hong Kong, and Singapore.

Due to the low investment cost of cultivation and a high price and good market for fruits, dragon fruit has become widely cultivated in central and southern provinces.

Two species of dragon fruit, Hylocereus undatus and H. costaricensis, are commercially grown in Vietnam.

Figure 2.7 A dragon fruit plant (Hylocereus undatus) with many wilted flowers.

(http://upload.wikimedia.org/wikipedia/commons/0/07/Dragonfruit_plant.jpg)

Methods of pollination for the dragon fruit cacti species differ: Hylocereus undatus is self-pollinating, while H. costaricensis, sometimes referred to as H. polyrhizus, is entirely dependent on cross-pollination between clones (Weiss et al. 1994).

Apis mellifera were diurnal visitors on the flowers of both Hylocereus species in Israel, collecting pollen but not nectar. In another study with “yellow pitaya”, S. megalanthus, in

Israel, more viable seeds per fruit, higher weight per fruit, and higher total soluble solids

(TSS) content were obtained by hand pollination (self and cross pollination did not differ) than either by auto-self-pollination or bee pollination by bumble bees (Dag and Mizrahi

2005). In Mexico, the pollination biology of H. undatus was studied (Valiente-Banuet et al. 2007). Percentages of flowers that set fruits were 54%, 40%, and 100% in hand self- pollination, hand cross-pollination, and unmanipulated self-pollination, respectively.

Nectar-feeding bats, Leptonycteris curasoae and Choeronycteris mexicana, were nocturnal pollinators, whereas A. mellifera was the primary diurnal pollinator. Numbers of fruits were higher with nocturnal pollination (77%) than with diurnal pollination

(46%).

Mango: Mango, an evergreen fruit tree, is planted widely over Vietnam, especially in Tien Giang and Nha Trang provinces in the south and Son La province in the north. Mango provides both nectar and pollen for bees. During flowering (December and January), many beekeepers move their colonies to mango orchards to sustain the bees until other nectar sources become available (Fig. 2.8).

There is lack of agreement in the literature on the pollination requirements of mangos. Sturrock (1944, cited by McGregor 1976) reported that mango flowers are self- fertile, but Popenoe (1917, cited by McGregor 1976) stated that while mango is self- fertile, cross-pollination increases fruit set. McGregor concluded that cross-pollination between cultivars is not necessary, but to obtain good crops of fruit, pollinating insects are required to transfer the pollen from anthers to stigma between trees within the cultivar. Calliphorid flies were the main visitors in the study by Bhatia et al. (1995). Fruit set was increased by ~4% with open pollination in comparison to bagged flowers (Bhatia et al. 1995). The primary pollinators identified by Fajardo et al. (2008) in the Philippines

were stingless bees (Trigona biroi), calliphorids (e.g., Chrysomya spp.), syrphids (e.g.,

Eristalis spp.), and bees (A. cerana and A. mellifera). There was no significant difference between the fruit set in caged (2%) and uncaged (3%) inflorescences in the absence of honeybees. After the introduction of bee colonies, the fruit set from the uncaged inflorescences (41%) was significantly higher than that for caged inflorescences (0.7%)

(Fajardo et al. 2008).

Figure 2.8 A mango tree in full bloom. (http://123.25.71.107:82/hoidap/uploads/news/2011_09/xoai4.jpg)

In Israel, two blow fly species (Chrysomya albiceps and Lucilia sericata), A. mellifera, and the housefly, M. domestica, play a significant role in mango pollination in most orchards. With cv. “Keitt” mangoes, the yield of small caged trees (e.g., pollinator access prevented) was very low (1 kg per tree), while the yield of open-pollinated trees was 61 kg per tree (Dag and Gazit 2001). In Japan, pollination of mangoes by

bumblebees (Bombus ignites) was not as efficaceous as with A. mellifera and A. cerana japonica. While European and Japanese honeybee visitation resulted in a large number of fruit weighing >200 g each, bumblebee pollination yielded smaller mangoes weighing <

100 g each (Mizuno et al. 2007). The fruit-to-seed mass ratio was high for all these pollinators: 72%, 69%, and 59% for Japanese honeybees, European honeybees, and bumblebees, respectively. Honeybees promoted embryo and seed formation in the cv.

“Irwin” mango. All harvested fruits were seedless and smaller in the honeybee-excluded inflorescences than for the seeded fruits obtained as a result of honeybee pollination.

Thus, honeybee pollination increased the yield of mangos (Sasaki et al. 1998). Apis mellifera was the most frequent of 21 insect pollinators visiting mango flowers in the Sao

Francisco Submedio Valley, Brazil, and was considered the most efficient pollinator of the crop in that region (Siqueira et al. 2008).

Peanut: An annual peanut crop is grown throughout Vietnam, but in greater acreages in the central provinces of Thanh Hoa, Nghe An, and Ha Tinh (Fig. 2.9). Peanut crops are grown on ~255,000 ha, with production reaching 505,000 tonnes per year

(General Statistics Office of Vietnam 2008). The value of the peanut crop is estimated to be 300 million USD per year. Peanut pollination requirements differ among cultivars.

Some older varieties of peanut, such as cv. “Krinkle Pearl”, are regularly pollinated by bees and other insects. Halictid and megachilid bees were most abundant during the time of the day (morning hours) when peanut pollination usually occurs, while honeybee activity was distributed over the entire day (Hammons 1963, cited by McGregor 1976;

Hammons and Leuck 1966). In comparison to plots in which insects were excluded with cages, both honeybee-visited and open-pollinated plots in which 1 – 2% of the total floral

visitors were A. mellifera had significantly higher numbers of pods per plant and seeds per plant, larger pods, and greater seed weights (Rashad et al. 1979). Hybrid peanut varieties (e.g., “Yellow flower”, “Narrow leaf”, and “Krinkle leaf” genotypes) are highly cleistogamous (i.e., self-fertilization occurs within the closed flowers). When seeds from peanuts of these three hybrid varieties were grown inside cages containing A. mellifera, only 0.82% of seedlings represented hybrids between the different genotypes. This suggests that honeybees can be used as pollen vectors in growing hybrid peanuts (Moffett et al. 1986), but the rate of crossing is <1%. Blanche et al. (2006a) determined that for several modern peanut cultivars (e.g., “Chifley”, “NC – 7”, and “SO95R”) grown in

Australia, bees (A. mellifera, Megachile sp., Lipotriches spp.) did not contribute to peanut pollination, apparently because new varieties of peanut are no longer attractive to the bees.

Figure 2.9 A peanut plant in bloom. (http://www.baovecaytrong.com/image/kythuat/trongtrot/dauphongrabong.jpg)

Cashew: Cashews are planted in the southern provinces and are a good nectar source for honeybees and a high diversity of other insect species that visit their flowers

(Table 2.1; Fig. 2.10). Heard et al. (1990) determined that only three species were efficient pollinators of cashew: A. mellifera and two species of native bombyliid (Ligyra) flies. A study of foraging activities of visitors to cashew flowers in Karnataka, India, showed that halictids (Pseudapsis oxybeloides) and honeybees (A. cerana) were the most prominent pollinators (Sundararaju 2000). A comparison of the foraging behaviour and pollination efficiency of two bee species (Centris tarsata and A. mellifera) visiting flowers of wild cashew trees was conducted in Brazil. Centris tarsata collected pollen with higher pollen removal efficiency index and deposited on stigmas twice as many pollen grains than A. mellifera (Freitas 1995; Freitas 1996; Freitas and Paxton 1998). In addition, C. tarsata pollinated more flowers per time unit. However, there were no significant differences between the number of flowers pollinated during individual visits of each bee species and the efficiency of setting seed. A study in India, using bagging experiments to exclude insect visitors, indicated that bees (A. cerana, A. florea, A. dorsata, and Bombus spp.) are efficient pollinators of cashews, increasing fruit set (7.9% vs. 0%) (Bhattacharya 2004).

Holanda Neto et al. (2002) identified the roles of self-pollination vs. cross- pollination, as well as A. mellifera foraging, for cashew production in northeastern Brazil.

They demonstrated that both self-pollination and cross-pollination can result in fruits.

Honeybees displayed foraging behaviour on cashew flowers that was conducive to cross- pollination. When honeybees were introduced into plantations originating from cloned material, fruit yield did not increase, although cashews have been reported to depend on

insect pollination. It seems likely that cashew has a mechanism of selective abortion through which it discards self-pollinated fruits. Honeybees can contribute to increased fruit yield only when genetically diverse cashew trees grow in the same orchard (Holanda

Neto et al. 2002).

Figure 2.10 Cashew tree with flowers and immature fruits.

(http://viipip.com/images/newsdata/2009/11/20091126091727_viipipdotcom_118

5cay%20dieu.jpg)

Citrus: Citrus fruit trees – oranges, mandarins, lemons, grapefruits, tangerines, and pomelos – are grown both in plantations and small rural groves (Fig. 2.11).

Honeybees and stingless bees are the primary pollinators of Citrus spp. (McGregor 1976;

Bhatia et al. 1995). In China, the fruit set of oranges pollinated by honeybees was 21%

higher than in orange groves not pollinated by honeybees; in addition, oranges from well- pollinated trees tend to have thinner peels and a flatter shape (YaoChun 1993). In an experiment in Arizona, USA, the yield of lemon trees caged with bees was ~20% more than trees caged to exclude bees during the blooming period (Moffett and Rodney 1975).

In tangelo (cv. “Orlando”) trees in Arizona, fruit yield was 49% more when trees were caged with A. mellifera versus when they were excluded. Tangerine (cv. “Fairchild”) trees caged with honeybees yielded more fruit than trees caged to exclude bees (Moffett and Rodney 1973).

Figure 2.11 An orange tree in Ha Tinh Province with many mature fruits.

In a study of mandarin pollination in southern California (Chao 2005), bagged flowers (i.e., excluding insect pollinators) had very low fruit set (0 – 2% in “Marisol”).

For the few fruits produced, there were very few seeds per fruit, indicative of very low

fertilization rates due to self-incompatibility. In contrast, cross-pollination between two clementine mandarin cultivars (i.e., “Nules” and “Fina Sodea”) and a mandarin cultivar

(i.e., “Afourer”), as well as their reciprocal crosses, produced high fruit set (Chao 2005).

Larger fruit size was positively correlated with larger numbers of seeds. In contrast,

Sykes (2008) determined that in some cultivars of mandarins, pollen germinated on the stigmas, pistils contained pollen tubes, and seeded fruit developed without insect pollination, suggesting that those cultivars were able to auto-pollinate.

2.2.2. Crops that do not benefit from bee visitation

A number of crops grown in Vietnam that are considered staple foods for

Vietnamese derive no benefit from bee visitations (Table 2.3).

2.3. Situation of bee pollination of crops in Vietnam

In Vietnam, crop pollination is poorly understood by farmers. Vietnamese beekeepers, in contrast to beekeepers in developed countries, not only do not receive payment for pollination services, but they often give a few bottles of honey to the owners of the land where they place their bees. Most of the growers do not realize that yields of many of their crops are greatly increased by the pollinating activities of honeybees. Some believe that honeybees damage their crops by foraging so heavily that the flowers fall off the plants. In some cases, this may reflect the rapid pollination achieved by bees rather than damage to the flowers.

Table 2.3 Important Vietnamese crops that do not benefit from insect visitation in their cultivation.

Total Area in Price Total References Scientific Name (Common Name) production hectares (x (USD/ value, USD (tonnes x (* for column 1; ** for columns 2, 3,4, Pollination type 1000) tonne) (x 1000) 1000) and 5) Oryza sativa (Rice) *Crane and Walker (1984); **General 7,200 36,000 273 9,792,000 Self-pollinated, wind pollination Statistics Office of Vietnam (2008) Zea mays (Corn) *Crane and Walker (1984); **General 1,070 4,000 152 624,000 Wind pollination Statistics Office of Vietnam (2008) Ipomoea batatas (Sweet potato) *Crane and Walker (1984); **General 178 1,500 121 176,000 Self-sterile (cultivated from plant Statistics Office of Vietnam (2008)

62 cuttings) Manihot esculenta (Cassava, manioc) *Rogers (1965); **General Statistics 500 8,000 121 966,000 Insect pollination of flowers Office of Vietnam (2008) (cultivated from plant cuttings) Saccharum spp. (Sugar cane) *Crane and Walker (1984); **General 290 17,400 33 574,000 Wind pollination Statistics Office of Vietnam (2008) *Canadian Association of Professional Vitis vinifera, V. labrusca (Grape) 3.1 41 728 30,000 Apiculturists (1995); Klein et al. (2007) Wind pollination, self-fertile **Nguyen (2008) *Values of crops in column 4 and 5 were converted from VND to USD (1 USD = 16,485 VND, Source: Vietcombank 11 Dec. 2008). **The original values in columns 4 and 5 were given in Vietnamese Dong (VND); data in column 4 were collected from government websites as well as from unofficial websites.

It is difficult to demonstrate to farmers the importance of honeybees and other insects as pollinators of their crops because many other insect pollinators, including wild bees, flies, butterflies and moths, all may contribute to increased yields of crops. Even if they have a vague belief that insect visitation may be beneficial for pollination of their crops, they fail to discriminate between effective pollinator visits and visits to flowers by non-pollinating insects.

In Vietnam, there have been only a few studies on pollination of agricultural crops. In 2005, the Vietnam Bee Research and Development Center (BRDC), in collaboration with the Southern Fruit Tree Research Institute, assessed potentially adverse effects of transgenic cotton on flower visitors like honeybees in Vietnam.

Although they suggested that the Bt toxin may negatively affect honeybee colonies and other animals (e.g., civets) that feed on honeybees (Le Thi et al. 2008), other research

(Rose et al. 2007; Hofs et al. 2008; Biao et al. 2009) suggest that Bt crops (e.g., cotton and corn) have negligible effects on bees.

Many Vietnamese crops rely on pollination by honeybees and other pollinators.

Some growers pollinate crops on smaller subsistence plantings (e.g., cucumber) by hand; this method of pollination is highly labour intensive and therefore expensive. Other growers utilize mechanical pollination that focuses on shaking plants so that pollen falls onto the stigmas (i.e., sweet pepper). Some larger commercial vegetable operations use chemicals (β-Naphthyloxy acetic acid, a phytohormone) to stimulate development of fruits. Finally, some growers, with the misconception that there is a problem with insect pollination, opt to grow cultivars/varieties that are self-pollinated or parthenocarpic, even though they know that those pollinated by insects usually produce better fruits.

In the last 10 years, production of some perennial commercial crops has increased. For instance, the production of coffee increased from 427,400 tonnes in 1998 to 961,200 tonnes in 2007 (i.e., an increase of 225%). Similarly, the production of cashew nuts increased from 54,000 tonnes in 1998 to 301,900 tonnes in 2007 (i.e., a 559% increase) (General Statistics Office of Vietnam 2008). The population of Vietnam, 85.2 million in 2007, is growing by more than 1.2% per year (General Statistics Office of

Vietnam 2008). Both domestic and international demands for food are growing (VTC

2008). In addition, living conditions of the Vietnamese people have improved remarkably as the national economy has grown. With more money available, people tend to consume higher quality products. Pollination often increases the shape and quality of fruits. To satisfy the demand for these higher quality fruits and vegetables, Vietnam imports many agricultural products that rely on insect pollination from foreign countries (e.g., US and

China). Another way to help meet this demand would be to improve pollination, thereby increasing yield of crops and quality of crops grown in Vietnam.

Pollination for seed production is also important. In the delta regions of the Red

River and Mekong River, farmers and seed companies devote some land to planting crops for seed production (e.g., cauliflower, yellow and white mustard, cabbage, and kenaf). To produce a good quality seed with high germination ability, seed crops need to be fully pollinated.

There have been various attempts to estimate the contribution of honeybees to total agricultural production (Robinson et al. 1989; Morse and Calderone 2000; Gallai et al. 2008). Table 2.4 summarizes the values of 8 of the crops reviewed above for which there is information on the extent of their dependence on bee pollination. Some crops

such as litchi, melon, cucumber, coffee, and avocado, depend heavily on pollination by insects, primarily honeybees. Other crops, such as oranges or dragon fruit, are less dependent on insect pollination, while for some crops, like corn, rice, and sugar cane, insect pollination is not required at all. The table reveals that honeybees contribute over

50% of the value of these important selected crops. The total pollination contribution of those crops to the Vietnamese economy is ~$1 billion USD. Moreover, this simple analysis is incomplete because there are no data for the many crops grown in small subsistence plantings, as discussed above (section 2.1). Also, many crops for which values of yields due to insect pollination are unknown were omitted from this table.

In 2005, the total economic value of pollination worldwide was estimated to be

€153 billion; that represents 9.5% of the total value of world agricultural production used for human food (Gallai et al. 2008). Honeybees and other pollinators contribute greatly to crop yields, as reviewed by Klein et al. (2007). Under certain conditions, such as low temperatures or in areas devoid of natural vegetation, when other natural insect pollinators are present in small numbers, the importance of honeybees is greater (Partap et al. 2000).

Table 2.4 Estimation of the contribution of honeybees on some major crops produced in Vietnam

Total Price Total value % of yield due Income due to Area (ha production Common name (USD/ (x 1000 to honeybee honeybees References x 1000) (tonnes x tonne) USD) pollination (USD x 1000) 1000) Avocado Kalman (1979), cited by 2.7 40 182 7,280 ~50 3,640 Crane and Walker (1984) Coffee Raw and Free (1977); 506.4 961.2 1,456 1,400,000 > 50 >700,000 Roubik (2002) Cucumber 0.3 2.5 243 605 90 545 CAPA* (1995) Litchi 28,200 to Anonymous (1981) 87.3 596.2 121 72,325 39 to 62 44,840

Longan 98 578 425 245,420 30 73,625 Crane and Walker (1984) Melons and water CAPA (1995) 3.2 837 121 101,530 80 81,220 melons Pumpkin /Squash 0.9 9 91 820 60 490 CAPA (1995) Sesame 44.7 28 1,519 42,460 32 13,590 Anonymous (1981) Total 901,300 to 1,870,440 918,000 Percent of income ~50% due to honeybee

*CAPA: Canadian Association of Professional Apiculturists Value of crops in column 5, 6 and 8 was converted from VND to USD (1 USD = 16,485 VND, Source: Vietcombank 11 Dec 2008).

In this assessment, it is shown that honeybees contribute greatly to crop production in Vietnam. Compared to other agricultural inputs like fertilizers and insecticides, costs of bee colonies for pollination purposes in Vietnam are relatively low compared to the benefits of increased yields as a result of improved pollination. In addition, in situations in which native pollinators are present in very low numbers, provision of honeybees to supplement pollination could save growers of specific crops large amounts of money that would have been spent on labour for hand pollination (i.e., cucumber).

The Vietnamese population continues to grow dramatically and as a result intensive agricultural practices will have to be utilized to ensure enough food is available.

One negative consequence of agricultural intensification is that wild bee communities are negatively affected, thereby interrupting their stabilizing effects on pollination services both at local and landscape scales (Klein et al. 2007). One consequence of deforestation of natural habitats is that distances increase between populations of pollinators that persist in patches of forest and the crops that benefit from their pollinating activities. Not only do populations of pollinators decline, but those that persist are only able to reach those crops within their foraging range. This phenomenon has been studied and modeled for coffee growing regions of Indonesia. Priess et al. (2007) estimate that at current and estimated rates of deforestation, coffee yields will decrease within the next two decades by up to 18% and net revenues per hectare may decline by as much as 14%, depending on the extent and location of the forested areas that are modified by human activities.

Most information about the pollination requirements of crops comes from studies in developed countries or other tropical regions of the world. Those studies may not be

applicable to tropical conditions in Southeast Asia, especially in Vietnam, where most of the population consists of rural subsistence farmers with diversified agriculture in small landholdings. Under these conditions, pollination has not been properly evaluated.

Moreover, throughout Southeast Asia, rural farmers are confronted with deforestation and changing methods of agriculture using insecticides that often affect natural pollinators.

Pollination requirements of many crops are underestimated or simply not considered. The role of honeybees and other insects as pollinators is not fully recognized; therefore, they and their habitats are not protected. It would be valuable to conduct research to identify, in rural communities, the diversity of pollinators, the pollinators that favour specific crops, and the increases in yields that occur with enhanced pollination. Major crops, such as litchi, longan, rambutan, mango, jujube, sesame, and peanut, should be studied because they are generally not widely cultivated in developed countries and their pollination requirements are not fully known. Such studies will be very important to identify the contribution that insect communities make to the economy of tropical countries in Asia. They also are important to raise awareness of people who work in the field of agriculture, from government workers to beekeepers and growers, about the value of honeybees in rural agriculture.

CHAPTER 3. THE FLORAL BIOLOGY OF JUJUBE, ZIZIPHUS MAURITIANA

Abstract

The floral biology of three sweet cultivars of the jujube (Ziziphus mauritiana) was studied in orchards in Hanoi, Vietnam, during September and October, 2010. Longevity of a single flower was ~1.5 days. In all three cultivars, anthesis (the time that flowers open and become sexually functional) occurred from 16.00h to 19.00h. Flowers had two distinct sexual phases: a male phase lasting 3-6 h (pollen became accessible when the anthers dehisced at around 16:00h) followed by a female phase lasting 22+ h. The stigma was strongly receptive on the 2nd day after anthesis. Jujube flowers bagged to exclude insects contained an average of 3.70 µl of nectar in the morning; despite continued nectar secretion, the volume declined to 1.47µl by mid afternoon, probably due to evaporation of water. Sugar concentration of nectar showed an inverse relationship, changing from

12% sugar in the morning to 65% sugar by mid afternoon. This information on floral biology is helpful in understanding the pollination biology of jujube.

3.1. Introduction

Jujubes grown commercially in Asia mostly belong to two species in the family

Rhamnaceae: Ziziphus mauritiana (commonly referred to as Ber in India) and Z. jujuba

(Liu and Zhao 2009). In Vietnam, only Z. mauritiana is cultivated and unless otherwise noted below, the following information relates to that species. Jujubes are fruiting trees with relatively large cultural and economic value in Vietnam (Nguyen 2008). Despite producing huge numbers of flowers, fruit yields are quite modest in Vietnam. The reasons

for these relatively low yields are unknown, but could relate to how they are cultivated or to the numbers of flowers that become effectively pollinated and develop into fruits. As the first step in understanding more about these factors that affect fruit yields, I studied jujube floral biology and how it relates to pollination of flowers (Chapter 4).

Jujube trees are evergreen shrubs or small trees that consist of several stems with drooping branches, many tiny thorns, and clusters of flowers at nodes of branches. The following description is from my personal observations in Vietnam. Plants are somewhat spherical in shape and have erect branches near the top and drooping branches near the base. Jujube trees are trimmed to maintain their height below ~5m and are very dense.

They generally occupy an area of approximately 9m2 (Fig. 3.1). The trunk is 10cm to

30cm diam depending on age of the rootstock. However, through regular pruning the fruit-bearing stems are considerably smaller in diameter. Jujube leaves are alternate and ovate, about 6-8cm long and 4 – 6cm wide, with three veins at the base. Floral buds are very small, with a diameter of 4mm; fully open flowers have a diameter of 6 – 7mm.

They are white or pale green with 5 sepals, 5 petals, and 5 stamens (Tel-Zur and

Schneider 2009). The fruit varies in shape and size. It can be round, oval or obovate, and

3 – 6cm long depending on the cultivar.

Figure 3.1 A jujube tree. The tree has erect branches near the top and drooping branches near the bottom, with a myriad of tiny flowers in clusters at nodes along branches.

Jujube can be propagated from seeds. However, usually relatively small branches

(referred to as scions) are cut from the desired cultivars and inserted into cuts made on a rootstock of an established jujube tree that has been cut off near the base. Grafting onto the base of an established rootstock improves tree vigour and adaptability of the new tree to local climatic and soil conditions. Most commercial jujube trees result from scions of sweet cultivars grafted onto rootstocks of a sour jujube cultivar. Usually 5 – 6 scions from a single source tree are grafted onto a rootstock (Nguyen 2008).

The typical time of year for planting jujubes in northern Vietnam is in November or December. Before planting, holes spaced 4 – 5m apart are dug, and 20 – 30kg of manure (pig, chicken, buffalo, and/or cow) is put into each hole. Jujube trees are planted

into the holes; then soil is put on top of the holes; water is provided and straw is placed around the tree base to retain moisture. About a month after planting, each plant is fertilized with 0.2 – 1.0 kg of N-P-K (N: 16, P: 16, and K: 8). Komix 201S® (a fertilizer that contains N: 3.5, P: 7, K: 2.3, Mn: 100ppm, Zn: 200ppm, Cu: 100ppm, and Mg:

500ppm) is sprayed onto leaves to facilitate fruit development (in September or October), with the usual dose being 6ml per tree (Ministry of Agriculture and Rural Development of Vietnam 2009; Dung 2009; Nguyen, S.V., pers. comm., 20129).

Jujubes require relatively high soil moisture, especially when they are nourishing developing fruits. Usually, jujube growers provide water 3 – 4 times during the periods when fruits are maturing (October to December). To obtain a high yield of good quality fruits, jujubes also require yearly pruning so that the trees redevelop every year (Fig. 3.2).

Pruning starts after harvesting season and depends on cultivar: normally in January for sour cultivars and February to March for sweet ones (Nguyen 2008).

From the literature, the period of time for bud development is reported to be 21 –

22 days depending on cultivar (Dhaliwal and Bal 1998). Floral buds appear in clusters at nodes along branches. They emerge in “axils of leaves of mature as well as current season shoots” (Desai et al. 1986, Garhwal 1997; cited by Pareek et al. 2007). The number of flowers in a cluster varies by cultivar: 9.5 flowers on average for cultivar

“Bomple” in Thailand (Wanichkul and Noppapun 2009) and 16-28 or 10 – 14 flowers in two different cultivars grown under different environmental conditions in India (Josan et al. 1980 and Garhwal 1997; cited by Pareek et al. 2007).

9 Nguyen, S.V., Hanoi, Vietnam (+84 904 629 563).

Figure 3.2 Jujube rootstocks, each with several side trunks that have grown from grafted scions. After collecting fruits, the trees are pruned. Intercropping with other crops (e.g., corn and squash) is typical.

Jujube flowers are reported to open at various times of the day, depending on species and cultivar. Zietsman and Botha (1992) reported that in Ziziphus mucronata, flowers opened throughout the morning and could be divided up into three sexual phases: an asexual phase when flowers start to open, followed by a male phase defined by fully opened flowers, and finally a female phase defined by the stigma being fully developed and receptive. Lyrene (1983) found that flowers from several cultivars of Z. jujuba opened in the morning while those from other cultivars of the same species opened in the afternoon. Pareek et al. (2007) reported that anthers dehisced about 2h after opening of flowers and pollen was available for no more than 4h (Dhaliwal and Bal 1998).

Longevity of a single flower was reported to be ~26h for Z. mucronala (Zietsman and

Botha 1992) to 33h for Z. jujuba (Ting et al. 2010).

Information on stigma receptivity in Z. mauritiana varies. Dhaliwal and Bal

(1998) reported that the stigma became receptive when it was shiny and light green, and peak receptivity occurred at flower opening. Singh et al. (1970) and Josan et al. (1981) reported that it was “receptive on the day of anthesis” (cited by Dhaliwal and Bal, 1998).

In contrast, others have noted that the stigma was most receptive on the day after anthesis as determined by visual observations and controlled pollination (Neeraja et al. 1996), and remained receptive for 13 – 24h (Desai et al. 1986, cited by Pareek et al. 2007) and in some cases up to 48h after flower opening (Godara 1980a, cited by Pareek et al. 2007).

There are several cultivars of jujubes grown in Vietnam. The sour cultivar (“Gia

Loc”) blooms in July – August and the sweet cultivars (i.e., F12, F8, and “Dao Muon”) bloom from late August to November. Among those cultivars, the sweet cultivars provide more nectar than the sour cultivar (Pham, T.V., pers. comm., 200910). However, there is no information on the floral biology and pollination of jujubes in Vietnam. Because of the paucity of information about jujubes in the literature in general and variabe reports about timing of anthesis and stigma receptivity, and numbers of flowers per cluster, the floral biology of jujube in Vietnam warrants further study. The aims of this study were to quantify the number of flowers per cluster, determine the timing of anthesis, determine how sexual phase of flowers changes with floral age, including the timing of anther dehiscence and stigma receptivity, and quantify the volume and concentration of nectar

10 Pham, T.V., Hanoi, Vietnam (+84 433 874 642).

and amount of sugar secreted by flowers. This information will aid in understanding the pollination of jujubes in Vietnam (Chapter 4).

3.2. Materials and Methods

3.2.1. Study site and cultivars for study

This study was carried out in Ha Dong district, 20km west of Hanoi (21.033N,

105.867E) between 20 August and 20 October, 2010 (Sunrise: 5:44h, Sunset: 18:00h on

September 15th). The vegetation in this area is similar to that of longan-growing areas and includes commercial acreage of papaya (Carica papaya), pumpkin and squash

(Cucurbita spp.), luffa (Luffa cylindrical), longan (Dimocarpus longan), litchi (Litchi chinensis), Citrus spp., maize (Zea mays L.), vegetables (Brassica spp.), and Chinese pea/cowpea (Vigna unguiculata). The climate in the region is classified as “Hanoi” – described as having a mean temperature of 23.8oC, relative humidity (RH) of 79%, and rainfall of 1700 – 1800mm (Ministry of Science and Technology 2009). During the experiments, the average daytime temperature recorded at the study site at 3h intervals between 06:00h and 18:00h was 28.7oC (the daytime range was 23 – 36oC) and relative humidity was 58.2% (range was 22 – 78%) (Pham, H.D., unpublished data).

Three sweet cultivars of jujube were selected for the study: F8, F12, and “Dao

Muon” (DM; meaning “late harvest”). All three were developed, selected, and named in

Vietnam. The F8 and F12 cultivars share several characteristics such as opening time of flowers (see below) and fruit weight. The average fruit weight (mean ± SE) at harvest of the F8 and F12 cultivars is 21.1 ± 0.26 g (n = 3) and 22.7 ± 0.52 g (n = 3), respectively.

The “Dao Muon” cultivar has a fruit harvest weight of 28.9 ± 1.29 g per fruit (n = 3;

Pham, H.D., unpublished data). F8 and F12 are harvested from December to February;

“Dao Muon” from January to March. Each of the cultivars was studied in two orchards and all blocks of trees were within 50 – 300m of each other. Within each orchard, the number of trees studied varied depending on the particular study.

3.2.2. Flowering characteristics and quantity of flowers per cluster

The flowering characteristics were observed and noted on several aspects: order of flower opening as it relates to position of floral clusters within the trees, distance between clusters on a branch, differences in timing of flower opening among trees in an orchard, and estimation of duration of flowering time between cultivars.

To quantify the number of flowers in a cluster, 4 trees from each experimental orchard were selected. Four branches from each selected tree were marked and 3 flower clusters on each branch were tagged before the first flower had opened. Throughout the entire flowering period of the cluster, newly opened flowers were counted daily, after which they were removed to prevent duplicate counts. After standardizing these data so counts of the number of flowers opening each day correspond to the day within the flowering period (e.g., numbers on day 1, day 2, etc.), the mean change in numbers of flowers opening each day could be determined for each cultivar.

3.2.3. Anthesis and floral longevity

Anthesis was studied on 3 trees of each cultivar in each orchard (3 trees x 2 orchards x 3 cultivars). Three branches were tagged for reference in the morning of 3

September, 2010. On each branch, ten floral buds that would open that afternoon (i.e., they were large and a slit had appeared at the bud tip) were randomly selected and marked

by tying short pieces of thread around the pedicel of each flower. To record the time of flower opening, all marked flowers were observed every 2h, from 06:00h to 18:00h.

Once flowers opened, changes in floral morphology were recorded from noon of the day of anthesis until evening of the next day. For this particular study, I selected 4 trees of each cultivar and repeated the following observations on 23 and 28 September,

2010. When the first flowers began to open on a tree, one branch per tree was tagged and ten floral buds that would open within the next few hours were randomly selected and marked with small pieces of string tied around the pedicel of each flower. Positions and conditions of the petals, stamens, anthers, and stigma were closely observed at 30 minute intervals, from 12:00h to 18:30h of the day of anthesis. On the following day, those flowers were observed until the stamens were fully recurved and the stigmatic surface was stained and/or had changed colour. Longevity of each flower was determined by recording the time elapsed from when a flower started to open (anthesis) until the stigma was dull and stained (Blair and Wolfe 2007). All flowers were exposed and could be visited by pollinators.

3.2.4. Pollen availability

Pollen availability was determined by visual observation during the afternoon of the day of anthesis because preliminary observations of flowers of all 3 cultivars on 5 and

6 September, 2010 indicated that anthers dehisced between ~1500h to 1800h.

Consequently, anthers of flowers that had recently opened were gently touched with a finger at 30 minutes intervals beginning at 15:00h. Pollen that stuck to the fingertip was evidence that anthers had dehisced during the preceding 30 minutes. This procedure was applied to ten flowers on four trees from each of the three cultivars on 15 September,

2010 (i.e., trees in only one orchard were studied). Because of the potential that touching the flowers with fingers affected the dehiscence of pollen, different flowers were tested at each time interval.

3.2.5. Stigma receptivity

The timing of stigma receptivity was determined three times, on 23, 28, and 29

September, 2010. On each date, 20 – 30+ flowers on four trees of each of the three cultivars were marked with string on the morning that they began to open. To estimate stigma receptivity, 5 marked flowers per tree were picked and placed facing upwards in a

8 cm Petri dish at the following times: Day 1 (day of anthesis), at 19:00h; Day 2: morning

(7:00h – 12:00h) and afternoon (13:00h – 17:00h); Day 3: afternoon (13:00h – 17:00h).

(For the first trial initiated on 23 September, 5 flowers per tree were also picked and tested at 16:00h and at 22:00h on the day of anthesis). A small drop (~0.025ml) of 3% hydrogen peroxide (H2O2) solution was placed directly on each flower to test for peroxidise activity which is highly correlated with stigma receptivity (Kearns and Inouye

1993). The flowers were observed under a dissecting microscope (8x) and the presence and approximate number of bubbles at the stigma were recorded ~30 sec after the H2O2 was added. The flower sampling interval during which maximum bubble production was observed indicated the peak of receptivity (Kearns and Inouye 1993).

Stigma receptivity on the 3 days of flowering was also tested with floral bagging experiments. Four trees from each cultivar were selected randomly. Four floral branches of approximately the same size on each tree were selected and tagged. Twelve flowers of the same age on each floral branch were marked with string. The four floral branches received different bagging treatments: (1) exposure of flowers to all types of pollination

for all 3 days flowers were open (open pollination; experimental control); (2) flowers exposed for the day of anthesis (Day 1), then bagged at the end of the day (Day 1 treatment); (3) flowers bagged right after marking floral buds, then exposed at the end of

Day 1 and bagged again at the end of the second day (Day 2 treatment); and (4) flowers bagged after marking floral buds, then exposed at the end of the Day 2 and bagged again at the end of the third day (Day 3 treatment). All bags were removed on Day 4 and immature fruits (those with a green slightly swollen ovary) were counted on Day 7. This procedure was repeated two times during the flowering season, on 27 September and 7

October, 2010. The day of peak receptivity was inferred from differences in initial incipient fruit set.

Temperature and RH were recorded during the study at regular intervals during the day (06:00h, 09:00h, 12:00h, 15:00h, and 18:00h) between 2 September and 14

October, 2010. Weather conditions did not vary greatly during this study. It was generally hot and humid, with light winds (1 – 5km per h) and a mixture of sun and clouds; mean temperatures were 23.0oC at 06:00h and 36.0oC at 12:00h. There were light rain showers about every 4 to 5 days. Flower visitors (observed in a companion pollination study; see

Chapter 4) did not differ during these experiments.

3.2.6. Nectar volume and concentration

Nectar was collected from the floral disc (the “base of the sexual column inside the circle of petals”; McGregor 1976) of flowers of all jujube cultivars in both orchards.

Three trees were randomly selected from each orchard and 4 floral branches were tagged from each tree. Preliminary observations indicated that only Day 2 flowers (those in female phase, on the day after anthesis) contained nectar. By bagging floral branches in

fine mesh netting in the afternoon (between 15:00h – 17:00h) of the day of anthesis, before flowers secreted any nectar, insect visitors were excluded. The next day (Day 2) the standing crop of nectar was extracted from 6 flowers during four time periods during the day. The nectar was extracted with 50µl Drummond micropipettes when the tip was placed on the flower disc. Because the process of collecting nectar may have harmed the flowers, nectar was collected from each flower only once. Sampling times were 08:00h –

10:00h, 10:00h – 12:00h, 13:30h – 16:00h, and 16:00h – 18:00h. These measurements were repeated on 4 days: 20, 27, and 29 September, and 1 October, 2010. Nectar was pooled from all 6 flowers sampled because the flowers were very small and contained small volumes of nectar. The volume of the pooled nectar was determined by measuring the length of the nectar column in the micropipette. Knowing length of the nectar column and the distance to the 50µl mark on the micropipette, the nectar volume could be calculated. The volume of the pooled nectar sample was then divided by 6 to obtain the average volume per flower. Sugar concentration was determined by placing the pooled nectar sample on the prism of a hand refractometer (Atago® Model # 2351, for 0 – 53% dissolved solids; Chinese refractometer, model unknown, for 0 – 90% dissolved solids) and reading the percentage of sugar to the nearest 0.5% (Kearns and Inouye 1993).

To quantify changes in mass of nectar sugar, volumes of nectar were multiplied by sugar concentrations (mg of sugar per μl of nectar). The sugar concentration in 1μl of nectar was obtained by converting the sugar concentration quantified with the refractometer to mg of sugars present in 1μl of nectar using the conversion table for sugar concentrations provided on the refractometer (Dafni et al. 2005).

3.2.7. Statistical analyses

Statistical analyses were conducted using SAS version 9.1 (SAS Institute Inc.,

Cary, NC) with statistical significance set at α = 0.05. Analyses of variance were used to analyze data on the number of flowers in a cluster and volume and sugar concentration of nectar collected. Comparisons of means were conducted with Tukey’s adjustment from the Mixed procedure in SAS. Correlation analysis was used to examine the relationship between nectar volume and sugar concentration. Boxplots are of standard construction, with the mean indicated by a cross (Pagano and Gauvreau 1993).

3.3. Results

3.3.1. Flowering characteristics and quantity of flowers per cluster

Jujube inflorescences consist of clusters of numerous flowers located at each node of a branch. Each cluster was 1.5 – 3.0cm from neighbouring clusters (Fig. 3.3). Within a cluster, when the first flower opened, the bud of the last flower that would eventually develop was not yet visible. Flowers could open first in any cluster along a branch (distal, proximal, or intermediate position), but it did seem that flowers in clusters that were more exposed to light opened before those that were more shaded (personal observation).

Within an orchard, a few trees initiated flowering earlier than others. The entire period of flowering for jujube in Ha Dong, Hanoi spanned ~75 days: from mid August to end of

October for F8 and F12 cultivars and from early September to mid November for the

“Dao Muon” cultivar.

Figure 3.3 Inflorescences of jujube (F12 cultivar). Flowers in various stages of development are evident within each cluster. The distance between each pair of adjacent clusters was 1.5 – 3.0cm.

Duration of flowering within individual clusters differed between orchards in F12 and “Dao Muon” cultivars (F = 184.89; df = 5, 246; P < 0.0001) (Table 3.1). Fig. 3.4 depicts the pattern of flower opening in the different cultivars. Of the 6 orchards studied, cultivar “Dao Muon” in orchard “F” had significantly more days of flowering (~38 days), followed by cultivar “Dao Muon” in orchard “E” (~34 days) and cultivar F12 in orchard

“D” (~28 days). Cultivar F12 in orchard “C” (~21 days) and cultivar F8 in both orchards did not differ (19 – 20 days).

Table 3.1 Mean duration of jujube flowering within flower clusters and quantity of flowers in a cluster (n = 4 trees per orchard) in orchards near Hanoi, Vietnam, 2010.

Jujube Orchard Number of days Number of flowers cultivar (Mean ± SE) in a cluster (Mean ± SE)

F8 A 19.4 ± 0.22 d* 21.3 ± 0.66 d

F8 B 19.8 ± 0.22 d 22.8 ± 0.55 d

F12 C 20.6 ± 0.58 d 25.1 ± 0.92 cd

F12 D 27.6 ± 0.54 c 27.8 ± 0.86 c

“Dao Muon” E 33.9 ± 0.96 b 36.5 ± 1.27 b

“Dao Muon” F 37.8 ± 0.69 a 46.0 ± 1.34 a

*Within a column, means followed by the same letters are not significantly different (P > 0.05, Tukey’s test).

There was significant variation in the number of flowers per cluster between cultivars (Table 3.1; F = 80.87; df = 5, 15; P < 0.0001). The quantity of flowers per cluster did not differ between the two orchards of the F8 cultivar and the two orchards of

F12 cultivar (Table. 3.1). However, the “Dao Muon” cultivar had significantly more flowers in a cluster than the F8 and F12 cultivars and the number of flowers differed significantly between the two “Dao Muon” orchards.

Dao Muon E Dao Muon F F12C 2.5 F12D F8A F8B

1.5

0.5 Number of newly opened flowers per flowers openedNumber day newly of 0 0 5 10 15 20 25 30 35 40 45 Days after anthesis of first flower

Figure 3.4 Numbers of flowers that opened daily within a flower cluster, for three jujube cultivars, each in two different orchards (n = 48 clusters for each cultivar). Data were standardized so that the first flower in every cluster opened on Day 1.

3.3.2. Anthesis and longevity

Jujube flowers started to open shortly before noon and continued into the afternoon.

Overall, most flowers, irrespective of cultivar, opened between 12:00h and 14:00h (93.7%)

(Table 3.2). However, only 2 flowers (0.4%) opened earlier (between 10:00h and 12:00h) and a small proportion of flowers (5.9%) opened between 14:00h to 16:00h.

Sex of flowers could be divided into three phases (Table 3.3), similar to what was reported by Zietsman and Botha (1992) for Z. mucronata. Floral opening (non-sexual period) lasted for about four hours. In this period, when the flowers first started to open, the petals were invisible (Fig. 3.5). The petals gradually extended and spread apart.

Initially the petals covered the stamens and were upright and perpendicular to the base of

the flower. The end of the asexual period and the beginning of the male phase was marked by the dehiscence of pollen from the anthers at around 16:00h for the F8 and F12 trees, and at about 16:30 h in the “Dao Muon” cultivar trees.

Table 3.2 Timing of initiation of flower opening of the 3 jujube cultivars grown near Hanoi, Vietnam, 2010*.

Times of day

Cultivars Total (flowers) 10:00-12:00h 12:00-14:00h 14:00-16:00h

DM 180 0 (0%) 170 (94.4%) 10 (5.6%)

F8 180 0 (0%) 176 (97.8%) 4 (2.2%)

F12 180 2 (1.1%) 160 (88.9%) 18 (10%)

Total 540 2 (0.4%) 506 (93.7%) 32 (5.9%)

*Data are numbers of flowers recorded followed by the percentage of flowers that opened during three 2-hour time intervals.

The male phase lasted for ~3 – 8h in the F8 and F12 cultivar trees, and for ~3 – 6h in the “Dao Muon” cultivar. Towards the end of the male phase, at about 21:00h, the petals had fully emerged, tiny drops of nectar became visible on the nectar disc and the stigma had emerged and elongated. The female phase was determined to have begun when the stigma had split into two lobes at the tip and had become shiny (Fig. 3.6a), and started producing bubbles when in contact with H2O2. This symptom was most clearly seen in flowers of the F8 cultivar. Oxygen bubble production in presence of H2O2 began at night of Day 1 between 19:00h to 22:00h and lasted until the end of the flowering period. Testing with H2O2 on the first night, from 22:00h to 23:30h, showed that 35/60

(58.3%) stigmas produced bubbles (details in section 3.3.3). During this phase, the stigma was fully developed. Stamens had tilted to the periphery and eventually drooped below the calyx. This stamen movement took place at night between 21:00h on the day of anthesis to 05:00h on the day after anthesis. Stigmas turned brown in colour (Fig. 3.6b) by or during the night of Day 2 (127/240 = 53% of stigmas became brown by 18:00h) and 227/240 (94.6%) by 06:00h of Day 3. Flowers were considered no longer functional when the tip of the stigma was stained and had changed colour to brown or black; a flower did not look fresh (its petals and pedicel became milky); or the flower fell off the tree. However, a small proportion of flowers looked fresh and appeared to be developing into fruits whether their stigma was stained or changed colour or not. Given that most stigmas had become brown-stained by the middle of the second night, the female phase was judged to have lasted 22 – 24+ h (initiated by 19:00 – 24:00h on Day 1 and terminated by 18:00h to midnight on Day 2). Longevity of a single flower was ~36h.

3.3.3. Stigma receptivity

At the start of the male phase (Day 1, 16:00h) and during late male phase, when the style was elongating (Day 1, 18:00h), application of H202 to the developing stigma resulted in generation of no bubbles. The production of bubbles from the stigma started at

19:00h (in 6/60 flowers = 10%) and had increased by 22:00h (35/60 flowers = 58.3%).

By morning of the second day of flowering, most of the stigmas responded to application of H202 by producing many small bubbles, indicating they were strongly receptive (56/60

= 93.3% of flowers tested from 07:00h to 12:00h). Receptivity continued through the afternoon (13:00h – 17:00h) of Day 2, as detected by strong generation of bubbles from stigmas in contact with H2O2 (176/180 = 97.8%) (Fig. 3.7). There were slight differences

among the cultivars. For example, the stigmas of F8 and F12 cultivars were divided into two lobes (Fig. 3.5a) that had a larger surface area for bubble production than the stigmas of “Dao Muon” cultivar flowers (Fig. 3.5b). Also, all stigmas looked very fresh and shiny on Day 2 (Fig. 3.5a). Stigmas may have remained receptive on Day 3, as inferred from the slower appearance and smaller number of bubbles in the presence of H2O2. Even at the end of Day 3 when all stigmas had turned to brown or black, 78.3% (141/180 flowers) still emitted low amounts of O2 bubbles in the presence of H2O2, perhaps indicative of tissue degeneration rather than stigma receptivity.

Table 3.3 Phases of flowering in the three cultivars of jujube.

Cultivars

Phases of Timing F8 F12 “Dao Muon” flowering

Flower Starting time 12:00h – 13:00h 12:00h – 13:00h 13:00h – 14:00h opening- asexual period Ending time 16:00h – 17:00h 16:00h – 17:00h 16:30h – 17:30h

Starting phase 16:00h – 17:00h 16:00h – 17:00h 16:30h – 17:30h Male phase Ending phase 19:00h 19:00h 19:00h

Starting phase 19:00h 19:00h 19:00h Female phase 18:00h – 24:00h 18:00h – 24:00h 18:00h – 24:00h Ending phase on Day 2 on Day 2 on Day 2

Figure 3.5 Phases of floral development in jujube. Upper left: Asexual period: A flower starts to open; Upper right: the flower is partly opened; petals are upright; and stamens are covered by petals and anthers have not dehisced; Lower left: Male phase: anthesis has occurred; anthers have dehisced and pollen has been presented. Lower right: Female phase: Stamens drooped onto petals; stigma developed and receptive.

a b

Figure 3.6 (a) Shiny receptive stigma with two lobes in F8 cultivar; (b) brown stigma of “Dao Muon” cultivar flower that is no longer receptive. Note that the petals have shrivelled up and the stamens have drooped downward.

Figure 3.7 Test for stigma receptivity in jujube. Two flowers with several bubbles on top of the stigma within the drop of H2O2 that indicated the stigma was receptive.

Experiments involving bagging of inflorescences also provided information on stigma receptivity over the three days that jujube flowers were open. A small proportion

(~7%) of flowers exposed to pollinators only on the first day or third day of flowering initiated slight swelling of the ovary (Fig. 3.8). In contrast, a much higher percentage of flowers (~15%) that were exposed to pollinators on either Day 2 or for all three days started to exhibit swelling of the ovary (hereby referred to as “incipient fruits”) suggestive that they had been successfully pollinated. These data seem to confirm that the

stigmas were most receptive on the second day of flowering. There were significantly more incipient fruit set on Day 1 and Day 3 compared to open pollination and Day 2 pollination (F = 4.52; df = 3, 84; P = 0.006).

0.2 a

0.15

0.1 b b

0.05 Proportion of incipient fruits incipient of Proportion 0 Day 1 Day 2 Day 3 Open

Treatments

Figure 3.8 Proportion of incipient fruit sets from flowers exposed to pollinators on the three days of flowering. Fruit set data were recorded one week after flowers opened. Day 1: floral branches were exposed to pollinators only on the first day of flowering; Day 2: floral branches were exposed only during the second day of flowering; Day 3: floral branches were exposed only on the third day of flowering; Open: floral branches were exposed to pollinators throughout the entire period the flowers were open (n = 96). Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test).

3.3.4. Nectar in flowers

Nectar secretion began in the late afternoon (around 17:00h) of the day of anthesis.

Nectar was initially secreted in a small amount from a disc-shaped floral nectary located around the central gynoecium (Fig. 3.6b). By the evening of Day 1, a small droplet of nectar was visible on the nectary. Flowers were bagged in the afternoon of Day 1 and the standing crop of nectar was collected and measured at several time intervals on Day 2 (Fig. 3.9).

Nectar

Figure 3.9 Nectar was withdrawn from a flower with a micropipette.

Nectar volume was highest in the morning and did not differ between the two morning collection periods (Figs. 3.10a, 3.11a, 3.12a, and 3.13a); it was lowest in the early afternoon between 13:30h – 15:30h. As nectar volume decreased, the sugar concentration increased (Figs. 3.10b, 3.11b, 3.12b, and 3.13b). There was a significant negative correlation between nectar volume and sugar concentration on each day (20

Sept.: r = − 0.77, p < 0.0001; 27 Sept.: r = − 0.83, p < 0.0001; 29 Sept.:, r = − 0.66, p <

0.0001; and 1 Oct.: r = − 0.82, p < 0.0001).

Figure 3.10 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 20 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 12:00h, 15:00h, and 18:00h were 33.5°C and 41.5%, 33.5°C and 40%, and 29.0°C and 71% respectively (n = 18 for each time point sample).

Figure 3.11 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 27 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 75%, 29.0°C and 59%, 33.5°C and 40%, 30.0°C and 53%, and 27.0°C and 71%, respectively (n = 18 for each time point sample).

Figure 3.12 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 29 September, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 75%, 30.0°C and 52%, 32.0°C and 38%, 30.5°C and 46%, and 26.0°C and 72.5%, respectively (n = 18 for each time point sample).

Figure 3.13 Nectar volume (µl) (a) and percentage of sugar (b) collected at different times of day on 1 October, 2010 from the three jujube cultivars. Temperature and relative humidity on that day at 6:00h, 9:00h, 12:00h, 15:00h, and 18:00h were 25.0°C and 76%, 26.5°C and 67%, 30.0°C and 48%, 30.5°C and 51%, and 28.0°C and 70%, respectively (n = 18 for each time point sample).

Nectar sugar increased over the course of the day. It was lowest in the early morning and highest in the afternoon (Fig. 3.14). There were weak interactions in the average amount of sugar per flower between different times of day and cultivars (F =

2.32, df = 6, 258; P = 0.034).

Dao Muon F8 F12

1.8

a 1.6 ac ab 1.4 abc abc abc 1.2 bd cd bc 1 0.8 de e e 0.6 0.4

0.2 Nectar sugar per flower flower per (mg)sugarNectar 0 8:00-10:00h 10:00-12:00h 13:30-15:30h 16:00-17:30h Observation intervals

Figure 3.14 Mean amount of nectar sugar (mg) per flower at different times of day (n = 24 for each bar) on the second day of anthesis. Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test).

3.4. Discussion

In this study, the three jujube cultivars differed from each other in the number of flowers per cluster, with the “Dao Muon” cultivar having more flowers per cluster (~36 –

47 on average) than the F12 cultivar (~26.5) and the F8 cultivar (~22). These values from

Vietnam are greater than reported by Wanichkul and Noppapun (2009) for the jujube

cultivar “Bomple” in Thailand (~9.5 flowers on average). Josan et al. (1980) and

Garhwal (1997) in India found that the number of flowers per cluster was 16 – 28 and 10

– 14, respectively (cited by Pareek et al. 2007), also lower than found in this study. In the current study, we randomly selected clusters of floral buds and then counted and removed all open flowers daily. This technique ensured that no flowers were missed, and may have contributed to larger counts than obtained in previous studies.

With trees of the F8 and F12 cultivars, the quantity of flowers in a cluster did not differ between orchards, but this was not true for “Dao Muon” cultivar trees. The difference found in number of flowers per cluster between the two “Dao Muon” orchards could be a consequence of any number of factors, including: irrigation regime; moisture retention in the soil; soil nutrients including N, P, K, Ca, Mg, and S); applications of fertilizers; and age of rootstocks. Older rootstocks are reported to produce more flowers, more fruits and larger fruits (Nguyen, S.V., pers. comm., 201211). However, although F8 trees had been cultivated for 9 years in one orchard and only 4 years in the other, the quantity of flowers within a cluster in those orchards did not differ. One possible explanation for the smaller number of flowers per cluster in one “Dao Muon” orchard is that those trees were re-grafted in 2010. For the first year after grafting, jujube trees are reported to have fewer flowers and lower fruit yields (Nguyen, H.C., pers. comm.,

201212). Further studies are needed to better determine the influences of cultivation regimes on quantity of flowers per cluster and how that may influence fruit yields.

11 Nguyen, S.V., Hanoi, Vietnam (+84 904 629 563).

12 Nguyen, H.C., Hanoi, Vietnam (+84 169 804 1955).

Flowering in Ziziphus mauritiana can be divided into three distinct phases. The development of stamens and the dehiscence of anthers before the maturation of the stigma indicate that this species is protandrous (Bertin and Newman 1993). Also, at the beginning of the female phase when the stigma first may become receptive (as indicated by minor release of bubbles in the presence of H2O2), the stamens had drooped onto the petals and the anthers were brown and lacked pollen. This strongly suggests that autogamy is unlikely. The large numbers of bubbles produced by stigmas in the presence of H2O2 throughout the Day 2 suggest that jujube flowers probably need to be pollinated on the second day when the stigmas are most receptive. These data suggest that pollinating agents that move pollen between male phase (Day 1) and female phase (Day

2) flowers within a tree and/or between trees should be very important for fruit set in this species. The relatively long duration of stigmatic receptivity compared to the duration of pollen presentation may enhance the chances of a flower getting pollinated. In Chapter 4, the pollination of this species is explored in detail.

H2O2 is a chemical that degrades cells (Bouchard and Purdie 2011); however, when placed on the stigma of a flower, it can be used to estimate stigma receptivity

(Zietsman and Botha 1992). In the case of the tiny jujube flowers, this procedure was not possible while they were attached to trees. Instead, the flowers were picked and a drop of

H202 was placed on the flower disc. However, the presence of oxygen bubbles through peroxidise activity does not only indicate stigma receptivity; it reflects total activity including cellular degradation that would occur naturally during floral bloom (Zietsman and Botha 1992). Consequently, peroxidise activity also is indicative of flower senescence (Grover and Sinha 1985). This fact could explain my observation that 78.3%

stigmas still emitted O2 in the presence of H2O2 at the end of Day 3 when all stigmas had turned to brown or black.

In the floral bagging experiment, incipient fruit set was much greater when flowers were exposed to pollinators on Day 2 than on either Day 1 or Day 3. This outcome also suggests that the stigmas were most receptive on the day after anthesis.

However, the incipient fruit set from open pollination was intermediate between the

Day1/Day 3 results and the Day 2 results. This result reflects the high variability in these results; also the sample size for the open pollination treatment was reduced because data from one tree were lacking. In the companion study of jujube pollination (Chapter 4), incipient fruit set that may reflect pollination was shown to be a poor predictor of which fruits would ultimately develop into mature fruits.

Phases of flowering in jujube are summarized in Table 3.4. I defined the start of the male phase by the dehiscence of pollen that corresponded with the separation of petals from stamens and the availability of pollen to insects. In Ziziphus mucronata,

Zeitsman and Botha (1992) stated that the male phase started before pollen became accessible to insects, at a time when the petals were upright and surrounded the anthers. It is not clear from their description whether anthers in Z. mucronata dehisced during this upright-petal stage or after the petals diverged.

Table 3.4 Phases of flowering in jujube.

Days of Day 1 Day 2 Day 3 flowering

At the end of the day By 06:00h, 95% Flower Fresh (18:00h): 53% brown stigmas changed to appearance stigmas brown or black

Shiny stigma; Stamens drooped below Shiny stigma; stamens Changes of stamens drooping calyx; stigma highly Flowers Anther drooping onto petals; Slower appearance floral onto petals; receptive based on open dehiscence O bubbles appear at a of bubbles structures; O bubbles first 2 extensive appearance of 2 high rate appear (%) bubbles; 100 response of stigma to

H2O2 10% 58.3% 93.3% 97.8% No data 78.3%

Time of day 12 13 14 15 16 17 18 19 20 21 22 23 24 7 – 12h 13 – 17h 7 – 12h 13 – 17h (hrs)

Phases Asexual period Male phase Female phase

There was a strong negative relationship between nectar volume and sugar concentration of the standing crop of nectar in flowers, as is commonly observed in many species (Nicholson and Thornburg 2007). Fig. 3.15a and b depict the average values (data from four days combined) for volume and sugar concentration of the nectar removed from flowers. A modest volume of dilute nectar was present in flowers in the morning when it was cooler and relatively humidity was high. Nectar volume decreased during the hottest, least humid period of the day, while sugar concentration increased, likely reflecting evaporation of water. The small volumes of nectar and open cup-shape of the flowers facilitated evaporation of water from jujube nectar (Nicholson and Thornburg

2007). Because pollinators were excluded from these flowers, it was not possible to determine if nectar secretion continued through Day 2 from these measurements of nectar volume. For this reason, I calculated the amount of sugar present in the nectar over the course of the day (Fig. 3.14). Sugar mass significantly increased over time, showing that nectar is continuously secreted throughout Day 2 (F = 91.69, df = 3, 258; p < 0.0001) when stigmatic receptivity is highest. It seems unusual that even with continued nectar secretion, evaporation rates were sufficiently high to result in decreased nectar volumes by early afternoon. There remains a possibility that water was resorbed from floral nectar by the plant (Nicholson and Thornburg 2007). Interestingly, there were no differences in nectar sugar between jujube cultivars at each of the four times it was quantified (Fig. 3.14).

Honeybees visit jujube flowers much more frequently between 0900h and 1300h, after which their visits decline over the rest of the day (see Chapter 4). Their foraging pattern may reflect their collection of relatively dilute nectar that has accumulated in

101

a a

4 µl)

3 b c 2

Nectar volume ( volume Nectar 1

0 8:00-10:00h 10:00-12:00h 13:30-15:30h 16:00-17:30h

Observation intervals a)

80 a 70 b 60 50

Sugar conc. (%) conc. Sugar 40 c 30 d 20 10 0 8:00-10:00h 10:00-12:00h 13:30-15:30h 16:00-17:30h

b) Observation intervals

Figure 3.15 Mean (± SE) of nectar volume (a) and sugar concentration (b) per flower in jujube at different times of day (n=72 for each bar). Means headed by the same letters are not significantly different (P > 0.05, Tukey’s test).

flowers overnight. Although the flowers continue to secrete nectar through the afternoon, foraging may decline because the nectar has become highly concentrated and viscous.

102

Moreover, because of the huge number (1500+) of beehives near the jujube orchards, there may be intense competition for food among nectar collectors that encourages them to forage in the morning when, because of the larger volume of nectar, bees would need to visit fewer flowers to fill up their crops.

To summarize, Ziziphus mauritiana in Vietnam has two distinct sexual phases.

Anthers dehisce and release their pollen (anthesis) in the afternoon of the day that flowers open; stigmas become most receptive the following day when flowers are actively secreting nectar. Flowers are visited during the day by many insects of many families, particularly honeybees and flies (Chapter 4). Other studies (see Chapter 4) have demonstrated that these diurnal insects are responsible for the pollination that results in jujube fruits. Due to the separation in time of male and female flower phases, insects that collect both pollen (from male phase flowers) and nectar (from different flowers in female phase) such as bees are likely to be the most effective pollinators.

103

CHAPTER 4. POLLINATION AND FRUIT SET OF JUJUBE (ZIZIPHUS MAURITIANA)

Abstract

The pollination biology of 3 sweet cultivars of the tree fruit jujube (Ziziphus mauritiana) was studied in orchards near Hanoi, Vietnam, during August and October,

2010. The numbers and proportions of floral visitors changed over time. Observations and collections of insect visitors to jujube flowers made in August showed that approximately 80% of all visitors were flies (Diptera). In contrast, of the insect visitors visually enumerated (n = 2095) during September, 84% were honeybees, Apis cerana, with the remainder primarily flies. Honeybees visited flowers most frequently on high outer branches, less frequently on low outer branches, and infrequently on low inner branches. Other insects visited flowers on outer branches equally but more frequently than inner branches. In the first pollination trial early in the flowering period, with five different pollination treatments, no fruits were produced; fruit set increased in the 2nd

(0.17%) and 3rd (2.21%) pollination trials. Fruit set recorded one week after anthesis suggested that wind, insect and self- pollination may result in fruits, but 7 weeks after anthesis only open pollination (unbagged flowers) and diurnal pollination treatments that involved insect visitation yielded fruits. The honeybee, Apis cerana, was a critical pollinator, with 97% of jujube production estimated to be attributed to its floral visits.

The requirement for insect pollination highlights the need for jujube growers to balance control of pests with pesticide applications against the need to protect essential insect pollinators, especially honeybees.

104

4.1. Introduction

Jujubes (Ziziphus mauritiana) are tree fruits that are important in the culture of

Vietnamese people, in part because they mature at the time of the lunar New Year

(Nguyen 2008). The trees are visited by several species of insects, particularly honeybees

(A. cerana) and flies (Singh 1984). Jujube has a high abundance of flowers, only a small percentage of which produce fruits. A deficit of pollinators and poor cultivation practices may contribute to suboptimal fruit production. Knowledge of pollinators and pollination biology may be important for increasing jujube yields.

Jujubes are evergreen trees with drooping branches and many tiny thorns. The basal trunk is 10 – 30cm in diameter, depending on age of the rootstock. However, through grafting of branches onto the rootstock, fruit-bearing stems are considerably smaller in diameter, dense, and through pruning usually kept below ~5m in height. Jujube leaves are alternate and ovate, about 6 – 8cm long and 4 – 6cm wide, with 3 veins at the base. The flowers are small, only 6 – 8mm in diameter. They are white or pale green with 5 sepals, 5 petals, and 5 stamens (Tel-Zur and Schneider 2009). The fruit varies in shape and size. It can be round, oval or obovate, and 3 – 6cm long depending on the cultivar (Fig. 4.1).

Jujube flowers attract many insect visitors including honeybees (A. cerana; considered the primary insect pollinator), and several taxa of Diptera including

Physiphora spp. (Ulidiidae), house flies (Musca domestica), and other flies (Singh 1984,

Devi et al. 1989).

Jujubes have been reported to have several modes of pollination. Some cultivars of Z. jujuba (also sometimes incorrectly refererred to as Z. jujube or Z. zizyphus: Liu and

Zhao 2009) are self-pollinating and self-fertilizing. Self-pollination is reported to occur

105

Figure 4.1 A branch of jujube tree with fruits. in 88% of Z. jujuba cultivars in China, but cross-pollination can increase fruit set ratio up to 600% (Liu et al. 2009). Fruit that develops from self-fertilized flowers are usually smaller than normal and prone to drop prematurely (Ackerman 1961). In contrast, several cultivars of Z. mauritiana failed to set fruit without cross-pollination (Pareek et al. 2007).

Ziziphus mauritiana, referred to as “Ber” in India, is the jujube species cultivated in northern Vietnam. Trees are usually planted in November or December. Before planting, holes spaced 4 – 5m apart are dug, and 20 – 30kg of manure (e.g., pig, chicken, buffalo, and/or cow) is put into each hole. Jujube trees are planted into the holes which are then filled with soil; water is provided and the soil around each tree is covered with a straw mulch to retain moisture. About a month after planting, each plant is fertilized with

0.2 – 1.0 kg of fertilizer in the following proportions: 16N : 16P : 8K. During flowering

106

in September and October, ~6ml of the fertilizer Komix 201S® (3.5N : 7.0P : 2.3K plus

100ppm Mn, 200ppm Zn, 100ppm Cu, and 500ppm Mg) are sprayed onto leaves to facilitate fruit setting and development (Ministry of Agriculture and Rural Development of Vietnam 2009; Dung 2009; Nguyen, S.V., pers. comm., 201213).

Jujube has several coleopteran, lepidopteran and dipteran pests whose larvae eat flowers and fruits. Pesticides widely used to control these pests include Fastac® (alpha- cypermethrin; HAI Agrochem Joint Stock Company), Marshal® (carbosulfan; FMC®

USA), Suphu® (fipronil; Allied Development Corporation), and Tasodant® (chlorpyrifos- ethyl and permethrin; Allied Development Corporation). Jujube is grown on small plots by different growers and pesticide applications are rarely coordinated so that everything is applied at the same time. One grower may spray pesticides a few days to a few weeks after another (personal observation). Consequently, many bees and other insect visitors are killed over an extended period of time while they forage for food. If insects are important for pollination, this regime of pesticide applications may result in fruit yields that are below potential.

Several cultivars of jujubes are grown in Vietnam. The sour cultivar (“Gia Loc”) blooms in July-August and the sweet cultivars (F12, F8, and “Dao Muon” or DM) bloom later, from late August to November. Among those cultivars, the sweet cultivars are reported to provide more nectar than the sour cultivar (Pham, T.V., pers. comm., 200914).

However, there is no information on the floral visitors and pollination of jujubes in

13 Nguyen, S.V., Hanoi, Vietnam (+84 904 629 563).

14 Pham, T.V., Hanoi, Vietnam (+84 433 874 642).

107

Vietnam. The paucity of published information on jujube pollination, particularly from

Vietnam, coupled with contradictory reports about pollination requirements, suggest that study of jujube pollination in Vietnam is warranted. The objectives of the current study were to determine: (a) the extent to which jujube cultivars grown in Vietnam benefit from insect pollination; (b) the relative importance of insect pollinator species, and (c) the economic value of insect pollination on jujube fruit yield.

4.2. Materials and Methods

4.2.1. Study site and jujube cultivars

This study was carried out in Ha Dong district, 20km west of Hanoi (21.033N,

105.867E) from 20 August to 20 October, 2010 (Sunrise: 05:44h, Sunset: 18:00h on

September 15th). The vegetation in this area is similar to that of longan-growing areas and includes commercial acreage of papaya (Carica papaya), pumpkin and squash

(Cucurbita spp.), luffa (Luffa cylindrical), longan (Dimocarpus longan), litchi (Litchi chinensis), Citrus spp., maize (Zea mays L.), vegetables (Brassica spp.), and Chinese pea

(Vigna unguiculata). The climate in the region is classified as “T.P Ha Noi” (Hanoi City) with mean annual temperature of 24oC, RH of 79%, and annual rainfall of 1700 –

1800mm (Ministry of Science and Technology 2009). During the experiments, the average daytime temperature recorded at the study site was 28.7oC (23oC – 36oC) and the

RH was 58% (22% – 78%) (data obtained between 06:00h – 18:00h between 2

September and 14 October, 2010).

Three sweet cultivars of jujube were selected for the study: F8, F12, and “Dao

Muon” (DM). All three cultivars were developed, selected, and named in Vietnam. The

108

F8 and F12 cultivars share several characteristics such as time of flower opening (see chapter 3, section 3.3.2) and fruit weight. The average fruit weight (mean ± SE) at harvest of the F8 and F12 cultivars is 21.1 ± 0.26g (n = 3) and 22.7 ± 0.52g (n = 3), respectively, while the “Dao Muon” cultivar is 28.9 ± 1.29g per fruit (n = 3; Pham, H.D., unpublished data). F8 and F12 are harvested from December to February, but “Dao

Muon” is harvested later – January to March. Six orchards (blocks) of jujube trees within

50 to 300 m of each other were studied. Within each orchard, the number of trees studied varied depending on the particular experiment.

During the course of the study, many beekeepers placed their beehives in villages near the experimental trees, to obtain jujube honey. The number of A. cerana beehives in the area in August at the beginning of the flowering period to heve been <150, but was estimated to have increased to 1500 hives by mid-September.

4.2.2. Floral visitors

4.2.2.1. Determination of the community of flower-visiting insects

Six floral branches on each tree were tagged on eight trees in two orchards from each of the three jujube cultivars. Six floral branches on each tree were tagged. To determine the effect of floral position on insect visitation, two upper exposed floral branches, two lower outer branches, and two lower branches on the inner part of each tree were selected. All insects observed on each floral branch were counted and recorded by scan sampling at different times of day (Dawkins 2007). The scans took ~90 min to complete and were initiated at 07:00h, 09:00h, 11:00h, 13:00h, 15:00h, and 18:00h over the six orchards. By the end of the last scan of the day it was dark and observations were

109

completed with the aid of headlamps. The study was replicated on 4 days (17, 19, 25, and

28 September, 2010). Representatives of floral visiting insects (except honeybees that can be identified in the field) were caught by insect net each day after the floral scans occurred; they were preserved in Eppendorf tubes (2.0ml) in a freezer for later identification. All insects were identified to the family level; in the case of honeybees and some of the larger flies, they were identified to genus or species. In addition to detailed scan-sampling, behaviours of the most common visitors on flowers, such as how they land on flowers and move between flowers, were observed and recorded.

Duration of individual floral visits of the common visitors was recorded between

6:00 – 9:00h, 9:00 – 12:00h, 12:00 – 17:00h, and 17:00 – 19:00h. The duration of a visit was defined by the elapsed time between an insect landing on and leaving a flower. The number of visits per minute of a particular insect taxon was estimated by dividing 60 sec by the duration of an individual visit of that taxon. Travel time between flowers of the visitors was ignored in estimating this value.

4.2.2.2. Pollinator determination

To quantify the amount of jujube pollen on an insect’s body, each insect examined was placed in a separate, clean Eppendorf tube (2.0ml) with 0.5ml or 1.0 ml of

50% ethanol. The tube was vortexed three times for 12 sec each time. Two aliquots of

2µl each of the solution were withdrawn from the tube and loaded onto the left and right sides of a haemocytometer. The haemocytometer was covered by a glass slide and then observed with a microscope at 400X magnification. Pollen grains were counted from the four squares at the corners and one square at center of the haemocytometer, then summed and divided by five to obtain the mean count per square. This procedure was replicated

110

four times, with the tube containing the insect vortexed each time to ensure thorough mixing of pollen grains in the solution. The haemocytometer was cleaned and dried after each count.

Total quantity of pollen on a single insect was calculated as follows:

Tp = (N ×V1)/V2 where:

Tp: Total quantity of pollen grains

N: Average quantity of pollen grains on a large square of the haemocytometer

V1: Volume of ethanol (ml) (0.5ml or 1.0ml)

V2: Volume of a large square = 0.004µl

In the case of honeybees, pollen in their corbiculae was included in this measurement.

After removing pollen grains from their bodies, the insects were pinned for identification. Insect specimens were identified by Dr. Stephen A. Marshall, School of

Environmental Sciences, University of Guelph, Canada. Specimens are stored at the Bee

Research & Development Center in Hanoi, Vietnam.

Index of pollinating efficacy: An index of pollinating efficacy, I, was calculated for the 4 most common floral visitors by multiplying several different factors, as follows:

I = V × P × J × F where:

I: Index of pollinating efficacy of a pollinator

V: Percentage of total floral visits by the specific pollinator species

P: Percentage of the insects of that species that had pollen on their bodies

J: Percentage of all pollen on the body that was jujube pollen

111

F: Quantity of flowers that a pollinator visited per min (travel time searching for flowers was not considered). This index assumes equal contact between insect and stigma for all pollinator species.

4.2.3. Pollination treatments

4.2.3.1. Fruit set of jujube

An experiment on pollination of jujube was carried out in both orchards of each of the three cultivars from September to October, 2010. Within each orchard, eight trees were selected randomly (n = 48) and numbered. On each tree, five floral branches of the same approximate size were tagged, and on each floral branch, short threads were tied around the pedicels of about 12 flowers (5 – 16) that would open later that day. The five floral branches then received different bagging treatments (Blanche et al. 2006b) (Fig.

4.2): (1) exposure of flowers to all modes of pollination (open pollination; experimental control); (2) enclosed by coarsely woven cheese cloth bags with ~0.5 mm mesh size, with all anthers from marked flowers removed by forceps before the dehiscence of anthers

(this treatment allowed only pollination by wind); (3) enclosed in mosquito-netting bags that excluded all insects from 18:30h to 06:30h but open in the daytime (diurnal pollination); (4) enclosed in mosquito-netting bags from 06:30h to 18:30h, but open at night (nocturnal pollination); and (5) enclosed continuously in paper bag (PBS

International, Bag type: Non-Standard), allowing for self-pollination only (Fig. 4.2). Note that the bagging in treatments 3 and 4 did not prevent wind pollination or self-pollination.

All bags were applied on the morning of anthesis. This experiment was replicated three times during the flowering period (early: 10 – 13 Sept; mid: 24 – 27 Sept; and late: 5 – 7

112

Oct.). The “Dao Muon” cultivar was not studied in the last replication. Data on fruit set were collected twice. The first assessment (one week after the experiment was established) was on 19 – 22 Sept for the first replication, 1 – 4 Oct for the second replication, and 11 – 14 Oct for the third replication; the second assessment of fruit set was made on 5 Dec (~7 weeks after anthesis).

Figure 4.2 Pollination treatments applied to jujube flowers. Upper left: open pollination with flowers continuously exposed to wind and insects; upper right: cheese cloth bag to allow for wind pollination; lower left: mosquito netting bag applied at night or during the day to quantify diurnal or nocturnal pollination, respectively; lower right: paper bag (PBS International, Non-standard) to quantify self-pollination.

4.2.3.2. Influence of insect pollination on crop yield and economic efficacy

At harvest, fruits were collected from all eight trees within the two orchards of the

F8 cultivar and one F12 orchard. Yield (kg of fruit) per tree on harvest days was

113

recorded, then summed after harvest ended to provide data on yield per tree.

The proportion of the fruit crop attributable to floral visits by honeybees was calculated as Phb = Ihb × (Yc − Yn) where:

Phb: proportion of the fruit crop attributable to honeybees

Ihb : relative index of pollinating efficacy of honeybees

Yc: yield from open pollinated flowers (control group)

Yn: yield from flowers bagged during the day (nocturnal insect pollination +

wind pollination + self-pollination [treatment 4 above])

The portion of the gross income from sale of jujubes on a per tree basis attributable to honeybee pollination, Ihb, was estimated as:

Ihb = Phb × Yt× $/kg fruit where:

Yt : yield per tree.

4.2.4. Statistical analysis

To analyze floral visitors, data were modeled as a repeated measure generalized linear model using PROC GLIMMIX in SAS 9.2 (SAS Institute Inc. Cary, MA) with a binary link function and autoregressive 1 (AR (1)) correlation structure. Random effects include day, orchard nested within cultivar, and tree nested within orchard and cultivar.

Fixed effects include floral visitors, positions of floral branches on trees, time of day, and cultivar. All first order interaction terms were modeled. Pairwise comparisons between

114

simple effects were adjusted for experiment-wise error rates using the Tukey-Kramer adjustment factor.

To analyze the effects of bagging experiments on fruit set, data were modeled as a split plot generalized linear model using PROC GLIMMIX in SAS 9.2 (SAS Institute

Inc., Cary, MA) with a binary link function. Trees were defined to be whole plots where weeks (the week when immature fruits were scored) were defined to be split plots.

Random effects included the orchard nested within cultivar and tree nested within orchard and cultivar where the whole plot error was defined to be the random effect of treatment by tree interaction nested within orchard and cultivar. Fixed effects included replications of the experiment (second and third replications only because no fruits were produced in replication 1), treatment, week and cultivar. The covariate “number of flowers” was also included in the model. The pairwise comparisons between significant effects were adjusted for experiment-wise error rates using the Tukey-Kramer adjustment factor.

4.3. Results

4.3.1. Diversity, abundance and pollination efficiency of floral visitors

4.3.1.1. Diversity of floral visitors

The mix of floral visitors changed over time. Insects belonging to 13 families were captured on jujube flowers at the study sites (Table 4.1). In August, ~80% of the total insects observed on flowers were flies. Some fly taxa were extremely abundant and very important flower visitors, especially the larger eristaline flower flies that breed in wet organic material readily available in riparian areas typical of jujube plantations, and the ubiquitous Chrysomya blowflies that abound in both agrarian and urban

115

environments. Most of the other insect visitors recorded were honeybees (many Apis cerana and a few A. florea). Small acalypterate flies (e.g., Lonchaeidae and Milichiidae) were common but were not recorded.

Table 4.1 Insect visitors recorded on jujube flowers.

Order Family/subfamily Genera and Species Hymenoptera Apidae Apis cerana, A. florea Hymenoptera Tiphiidae Hymenoptera Vespidae: Vespinae, Polistinae, Vespa spp., Polistes spp., and and Eumeniinae others Diptera Syrphidae/mostly a Eristalinus megacephalus, few species of Eristalinae, but Eristalinus lathyrophthalinus also some species of Syrphinae Diptera Sarcophagidae Diptera Calliphoridae Mostly Chrysomya spp. Diptera Muscidae Musca spp. Diptera Rhiniidae Diptera Tachinidae Diptera Lonchaeidae Diptera Milichiidae Coleoptera Scarabaeidae Hemiptera Tessaratomidae Tessaratoma papillose

In contrast to the fly-dominated assemblage of floral visitors observed in August, honeybees dominated the observations and collections made in September. By then, more than 1500 colonies of A. cerana had been moved into nearby villages and of the total visitors to flowers visually enumerated (n = 2095), A. cerana was most frequent (84%).

The balance consisted of visits by Calliphoridae (especially Chrysomya spp.), Syrphidae

(especially two species of Eristalinus), other flies, other Hymenoptera (including a

116

modest number of A. florea bees), and a few Lepidoptera and Coleoptera. Visits by small acalyptrate flies were not recorded.

4.3.1.2. Floral visitor behaviour

In September, when detailed observations of pollinator behaviour were made, many more honeybees were observed on jujube flowers in comparison to non-honeybee visitors; moreover honeybee visits varied temporally (Fig. 4.3). Because of the large number of scans with no insects present, the data were analyzed logistically as presence or absence of either honeybees or all other visitors per inflorescence. There was a highly significant interaction between the class of floral visitor and time of day (F = 53.67; df =

5, 13738; P < 0.0001) (Fig. 4.3). Because of the significant interaction, odds ratios were calculated for each pairwise comparison of numbers of insect visitors from successive time periods (Table 4.2). The proportions of floral branches visited by A. cerana during scans differed between each pair of time intervals (P < 0.0001), except for the periods beginning at 09:00h and 11:00h (P = 0.763). Their activity peaked between 9:00h and

12:00h. Diurnal visits to jujube flowers by the other main taxa of insect visitors (data pooled) were less frequent and did not vary temporally (P > 0.05 for all pairwise comparisons).

Positions of floral branches also affected the presence of both A. cerana and other visitors (Fig. 4.4). Because there was an interaction between positions of floral branches and class of visitors (F = 9.9; df = 2, 13738; P < 0.0001), separate pairwise comparisons were made. The high outer branches received 9.79 times more visits by honeybees (95% confidence intervals of 6.29 – 15.223; P < 0.0001) than the low inner branches, and 1.9 times more visits (confidence interval: 1.465 – 2.477; P < 0.0001) than the low outer

117

branches; very few low inner branches had honeybees present. Low outer branches had

honeybees present 5.13 times more often than low inner branches (confidence interval:

3.29 – 8.00; P < 0.0001). Visits by insects other than honeybees to the two categories of

outer branches did not differ (P > 0.05); however, low inner branches were visited 3.47

times less frequently by non-honeybees than low outer branches (CI = 1.95 – 6.17;

P < 0.0001) (Fig. 4.4).

Table 4.2 Odds ratios and their 95% confidence limits for comparisons of the proportions of flowering branches of jujube with insect visitors between one time interval and the preceding time interval, during September, 2010.

Apis cerana Other insect visitors

Comparison of Times more 95% Times more 95% forager likely to visit P Confidence likely to visit P Confidence collection times (Odds ratio)* interval (Odds ratio) interval

09:00 to 07:00h 7.35 0.0001 4.37 – 12.35 0.66 0.7982 0.31 – 1.39

11:00 to 09:00h 1.21 0.7634 0.87 – 1.69 1.18 0.9999 0.55 – 12.54

13:00 to 11:00h 0.62 0.0001 0.45 – 0.86 1.11 1 0.53 – 2.33

15:00 to 13:00h 0.52 0.0001 0.36 – 0.75 1.35 0.955 0.68 – 2.66

18:00 to 15:00h 0.02 0.0001 0.01 – 0.08 0.67 0.9229 0.29 – 1.56

*The odds ratios reflect comparisons of data from one time interval to similar data from the preceding time interval; for example, the odds ratio for honeybees between the two times 09:00h-07:00h was 7.35, which indicates that honeybees were 7.35 times more likely to visit at 09:00h than at 07:00h.

118

0.9 Apis cerana

0.8 Other visitors 0.7

0.6

0.5

0.4 a ab 0.3 bc

0.2 d

0.1 e f Proportion of floral branches with visitors present visitors with branches of floral Proportion eg g g eg eg eg 0 07:00h 09:00h 11:00h 13:00h 15:00h 18:00h

Starting times of observation periods

Figure 4.3 The proportion of floral branches with Apis cerana or other insect visitors present at different times of day during September, 2010 (n = 1152 floral branches). The proportion of branches with honeybees present differed between all successive time intervals except for those beginning at 9:00h and 11:00h.The proportion of branches receiving visits by other insect species did not differ temporally (P > 0.05 for all pair wise comparison, Tukey’s tests). Data points are the proportion of floral branches with visitors present, with 95 percent confidence intervals, the same letters indicate no significant difference.

There was also a significant interaction between time of day and jujube cultivar in the proportion of floral branches receiving visits (Fig. 4.5) (F = 4.17; df = 10, 13738; P <

0.0001) and between the time of day and floral branch position (Fig. 4.6) (F = 4.76; df =

10, 13738; P < 0.0001). Odds ratios and their confidence intervals for these two comparisons are presented in Appendix 1 and 2. Relative to the other cultivars, flowering 119

0.7 a

0.6 Apis cerana

0.5 Other visitors

b 0.4

0.3

0.2 d d cd

0.1 e Proportion of floral branches with present branchesfloralofvisitors with Proportion 0 High outer Low outer Low inner

Positions of floral branches

Figure 4.4 The proportion of floral branches with Apis cerana or other visitors present on jujube floral branches in three different positions on trees. The presence of A. cerana differed with respect to each floral branch position. Visits by non-honeybees did not differ between high outer and low outer branches (p = 0.954). Data points are the estimates of proportion of branches on which visitors were observed during scan sampling, with 95 percent confidence intervals (n = 1152 scans of flowering branches in each branch position category). The same letters indicate no significant difference.

branches on “Dao Muon” cultivar trees were visited more often later in the day while visitors to F12 cultivar trees declined in late afternoon (Fig. 4.5). It seems that insects shifted their visitation from high outer to low outer floral branches and finally low inner branches as the day progressed (Fig. 4.6).

The four main insect taxa that visited jujube flowers differed in the duration of their visits (F = 5.25; df = 9, 1003; P < 0.0001; Fig. 4.7). For A. cerana and flower flies

120

Dao Muon F8 F12

0.4

0.35 a a a a a a a 0.3 a a ab

0.25 a 0.2 a a b 0.15 a

0.1 a a a 0.05

Proportion of floral branch with insect visitors with branchfloralof Proportion 07:00h 09:00h 11:00h 13:00h 15:00h 18:00h Starting times of observation periods

Figure 4.5 The proportion of floral branches of the 3 jujube cultivars (Dao Muon, F8 and F12) receiving visits by all species of insects at different times of day. Data points are the proportions of branches on which visitors were observed during scan sampling, with 95% confidence intervals (n = 384 scans of flowering branches in each cultivar/observation period). Within an observation period, the same letters indicate no significant difference.

(Syrphidae), the duration of each floral visit did not vary significantly with time of day, but the visits by honeybees were consistently shorter in duration than for any of the other taxa of flower visitors. Apis cerana spent an average of only 1.84 ± 0.550 sec visiting each flower, significantly shorter than visits observed for flower flies (4.17 ± 0.550 sec).

Blow flies and house flies remained considerably longer on individual flowers (mean durations of 5.03 ± 0.551 and 7.13 ± 0.560 sec, respectively). Blowflies remained on flowers significantly longer in mid-afternoon than in the two morning collection intervals. Houseflies showed a progressive increase in flower visit duration from morning to evening, with significantly longer visits by late afternoon compared to morning visits.

121

High outer Low outer Low inner

0.7

a 0.6

0.5 a a 0.4 b a a a a 0.3 ab a b 0.2 b bc c 0.1 b a a b

0 Proportion of floral branch with insect visitors with branchfloralof Proportion 07:00h 09:00h 11:00h 13:00h 15:00h 18:00h Starting times of observation periods

Figure 4.6 The proportion of floral branches at 3 different positions (high outer, low outer and low inner floral branches) receiving visits by all species of insects at different times of day. Data points are the proportions of branches on which visitors were observed during scan sampling, with 95% confidence intervals (n = 384 scans of flowering branches in each branch position/observation period). Within an observation period, the same letters indicate no significant difference.

Generally, all visitors tended to move from one flower to the next on the same floral branch. While foraging, all three of the main fly taxa landed on the flowers. They clung to floral parts (e.g., petals, stamens, sepals, pistil) while probing for nectar. Their forelegs and/or mid-legs touched dehisced anthers and pollen adhered to their legs. When they foraged on an adjacent flower, their legs frequently came in contact with the stigma.

After foraging on a flower, the flies often landed on a leaf and licked or groomed their legs (Fig. 4.8); following this behaviour, the pollen which may cling to their proboscis may also be deposited on the stigma of the next flower.

122

A. cerana Blow flies Flower flies House flies 10 40a

72ab 8 60bc 7 60bcde 72abd 51bcde 6 72cde 72cef 5

4 72egh 3 72defgh 72defg 60defgh

2 Mean duration of floral visits visits (sec)floralof Meanduration 72ghi 1 72hi 72i 60ghi 0 06:00-09:00h 09:00-12:00h 13:00-17:00h 17:00-19:00h

Observation intervals

Figure 4.7 Mean (± SE) duration of floral visit by honeybees (Apis cerana) and members of three fly families during each floral visit, divided into four temporal observation periods. The 4 main insect taxa that visited jujube flowers differed in the duration of their visits (P < 0.0001). For A. cerana and flower flies (Syrphidae), the duration of each floral visit did not vary significantly throughout the day. The visits by honeybees were consistently shorter than for any of the other taxa of flower visitors. Means with the same letters are not significantly different (P > 0.05, Tukey’s test). Numbers located before letters are numbers of insects studied.

Honeybees (A. cerana) landed either on an individual flower or, because of their larger size, on several adjacent flowers simultaneously (Fig. 4.8, lower left). While foraging for nectar and pollen, they contacted stigmas, anthers, and other parts of flowers with their legs and probosces. Because jujube flowers occur in clusters containing many flowers, honeybees flew rapidly between adjacent flowers and neighbouring flower clusters on the same tree or different trees.

123

Figure 4.8 Insects visiting jujube flowers. A blow fly (Calliphoridae) foraging for nectar (upper, left) and lapping its foreleg (upper, right). A honeybee, Apis cerana, standing on two adjacent flowers simultaneously (lower, left) and in contact with the anthers of a flower (lower, right).

4.3.1.3. Pollen loads of visitors

The primary visitors to jujube flowers were divided into four groups: honeybees

(A. cerana), Syrphidae (flower flies), Calliphoridae (blow flies), and Muscidae (house flies and their relatives). Among the floral visitors, flower flies had the highest number of jujube pollen grains on their bodies, followed by A. cerana (Table 4.3). Houseflies and blowflies had relatively small amounts of pollen. Nearly all the A. cerana sampled (91%)

124

had pollen on their bodies. Much smaller proportions of the non-syrphid flies had pollen

of any species on their bodies (8 – 33% depending on type of fly). Also, A. cerana, as is

typical of all honeybees, had a high degree of floral constancy and carried very little non-

jujube pollen on their bodies (~0.1%). Approximately 95% of the pollen on flower flies

was from jujube flowers.

Table 4.3 Percentage of insects that carried pollen and quantity of jujube pollen on specimens sampled on jujube flowers. (Values include corbicular loads of pollen carried by Apis cerana).

No. of insects that Mean No. of Mean No. of Percentage carried pollen pollen grains jujube pollen Foragers N of jujube per insect grains per insect N % pollen body (±SE) body (±SE)

Flower flies 319,500 ± 302,300 ± 33 26 78.8 94.6 (Syrphidae) 75,700 76,700

Apis cerana 227,500 ± 227,200 ± 64 58 90.6 99.9 (Apidae) 23,500 23,500

House flies 13,900 ± 13,900 ± 9 3 33.3 100 (Muscidae) 10,200 10,200

Blow flies 24 2 8.3 500 ± 360 500 ± 360 100 (Calliphoridae)

4.3.1.4. Index of pollinating efficacy

The relative pollinating efficacy of the four main taxa of pollinators was

estimated, based on the relative abundance of each type of insect on jujube flowers (data

obtained from scan sampling of floral branches), the amount of jujube pollen that the

125

visitors carried, and the duration of floral visits. A.cerana was by far the most important pollinator, estimated to have contributed 97% of all jujube pollination, followed by flower flies (Eristalinus megacephalus and E. lathyrophthalinus, 2%), blow flies (0.3%), and house flies (0.1%) (Table 4.4). Indices of pollination efficacy for A. florea and other insect visitors were not calculated due to their scarcity.

Table 4.4 Pollination efficacy of honeybees (Apis cerana) and other insect pollinators visiting jujube flowers.

Percentage Percentage Percentage Flowers Relative Pollination Pollinators of with of jujube visited pollinating index* pollinators pollen pollen per min efficacy**

Apis cerana 84 91 100 32.4 24.77 0.974

Flower flies 5 79 95 14.4 0.54 0.021 (Syrphidae)

Blow flies 8 8 100 12 0.08 0.003 (Calliphoridae)

House flies 1 33 100 8.4 0.03 0.001 (Muscidae)

*Pollination index is the product of the four variebles **The relative values, with relative pollination values for the 4 main taxa of pollinators summing to 1.

4.3.2. Effect of pollination on fruit set

Pollination experiments were replicated three times: early (Replication 1), midway (Replication 2) and late (Replication 3) in the flowering period. Overall, there was no significant difference in incipient fruit set (Week 1 data) between the three cultivars (F = 1.78; df = 2, 3; P = 0.310). However, there were significant differences in the number of incipient fruits at week 1 compared with the number of developing fruits at

126

week 7 (F = 41.21; df = 1, 542; P < 0.0001); in the number of fruits between Replicate 2 and 3 (F = 78.76; df = 1, 542; P < 0.0001); and between pollination treatments (F = 9.92; df = 4, 188; P < 0.0001).

The first replication (i.e., 2880 individually marked flowers from 48 trees of the three cultivars) yielded 12 incipient fruits at week 1, representing four of the pollination treatments (Table 4.5). Five days later, 10 of those incipient fruits were aborted; the last two of those fruits dropped over the next two weeks. No fruits remained by week 7.

In Replication 2 (2880 flowers on the same trees as in Replication 1), 18 incipient fruits were recorded at week 1 from three treatments: open pollination, diurnal pollination, and self-pollination treatments (Table 4.5). Incipient fruit set in those treatments did not differ (P = 0.083). Two treatments in the second replication (i.e., nocturnal pollination and wind pollination) set no fruits. Consequently, those treatments could not be included in the statistical models for analysis. By week 7, only 5 (0.2%) developing fruits remained: 4/576 (0.7%) of flowers from the open-pollination treatment and 1/576 (0.2%) flowers exposed to diurnal pollinators. The number of fruits obtained in those two treatments did not differ (P = 0.201).

In Replication 3 (i.e., 1806 flowers on 32 trees (F8 and F12 cultivars only) at week 1, incipient fruits were evident in all treatments and the effects of treatments were significant (P = 0.029); most pairwise comparisons between treatments gave P values >

0.05, except the comparison between diurnal pollination treatment and wind pollination treatment (P = 0.033, odds ratio = 5.431 with 95% confidence intervals of 1.096 – 26.91).

By the time of the second assessment of pollination success at week 7 after anthesis, the situation was very different. Extensive fruit abortion had occurred and similar to results

127

of Replication 2, only two treatments yielded 40 fruits (2.2%): open pollination (16 fruits) and diurnal pollination (flowers bagged at night; 24 fruits). The numbers of flowers that set fruits in these two treatments (Table 4.5) did not differ (P = 0.110).

Table 4.5 Effect of bagging experiments on fruit set. Data are numbers of developing fruits recorded, followed by the percentage of flowers studied that set fruits. There were no fruits remaining on trees by week 7 of Replication 1. Noctural and diurnal treatments include effects of wind and self-pollination).

Total number of developing Fruit (% flowers setting fruit)

Replication 1 Replication 2 Replication 3

Total Total Total Treatments Week 1 Week 1 Week 7 Week 1 Week 7 flowers flowers flowers

6 12 4 33 16 Open pollination 576 576 365 (1.0%) (2.1%) (0.7%) (9.0%) (4.4%)

Nocturnal 1 0 0 19 0 576 576 372 pollination (0.2%) (0%) (0%) (5.1%) (0%)

4 3 1 39 24 Diurnal pollination 576 576 338 (0.7%) (0.5%) (0.2%) (11.5%) (7.1%)

1 0 0 15 0 Wind pollination 576 576 370 (0.2%) (0%) (0%) (4.1%) (0%)

0 3 0 31 0 Self-pollination 576 576 361 (0%) (0.5%) (0%) (8.6%) (0%)

Treatments contained 5 floral branches on each tree. Total number of flowers used per treatment varied from 5 to 16 per floral branch with a mean of 11.71. Open pollination = unbagged flowers and all possible modes of pollination; nocturnal pollination = exposed to visitors from 18:30 to 6:30, plus possible effects of wind and self-pollination; diurnal pollination = exposed to visitors from 6:30 to 18:30, plus possible effects of wind and self-pollination; wind pollination = flowers with anthers removed and bagged continuously using cheese cloth to exclude all insect visitors; self-pollination = bagged continuously to exclude both wind and pollinators.

128

4.3.3. Jujube yields

Harvest of fruits started during the second week of December and finished near the end of March. The average yield for F8 cultivar trees was 126 ± 12.9kg per tree in orchard A and 46 ± 1.8kg per tree in orchard B (Table 4.6). The average yield for F12 in orchard C was 105 ± 6.4kg.

4.4. Discussion

The diversity of flower-visiting insects on jujube flowers was relatively high.

Captured specimens represented 10 families of insects, with the most abundant being

Apidae (A. cerana), Syrphidae (Eristalinus megacephalus and E. lathyrophthalinus), and

Calliphoridae (Chrysomya spp.). Honeybees were scarce in August, but by the time detailed visitation data were collected in September, beekeepers had moved large numbers of hives to the region and A. cerana dominated the community of floral visitors.

Nadia et al. (2007) recorded insect visitors to Ziziphus joazeiro flowers in Brazil; 77.5% of visits were by A. mellifera, followed by wasps (20.4%) and flies (2.1%). In India, Singh

(1984) observed 7 dipteran and 11 hymenopteran species on flowers of Z. mauritiana.

In this study, A. cerana spent only 1.84 sec on average at each flower, a significantly shorter amount of time than the other three major taxa of floral visitors.

Rapid floral visits, as well as the high proportion of honeybees with pollen and the high proportion of that pollen that was jujube pollen, resulted in A. cerana having a very high relative index of pollinating efficacy (0.974). Honeybees also had large numbers of jujube pollen grains on their bodies that contributed to their effectiveness as pollen vectors.

The next most important group of pollinators were the flower flies, predominantly

Eristalinus, with a relative pollinating index of 0.019. Most of the syrphids observed

129

were relatively large bodied and very hairy. More then 75% of them carried pollen, and they had the largest average amounts of pollen on their bodies, 95% of which was jujube pollen. The large amount of pollen on the bodies of these flies was surprising, but confirmed as being possible by Jeff Skevington (pers. comm.). The syrphids spent about

5 sec on each flower, 2.5 times longer than the honeybees. Blow flies and house flies were uncommon, tended to lack pollen on their bodies, and stayed on flowers for relatively long amounts of time, resulting in them having low relative pollinating indexes

(0.003 and 0.001, respectively).

The relative pollinating efficacy for A. florea and other insect visitors was not calculated due to insufficient data. Apis florea could be a significant pollinator of jujube in situations where they are more common. I observed them collecting pollen, indicating that they would have pollen grains stuck to their bodies as they moved between flowers.

In this study they constituted only 1.3% of all floral visits. For all insect taxa that visit jujube flowers, direct quantification of their feficiency of pollen deposition on stigmas is needed to fully evaluate their relative contribution to pollination and fruit production

(Primack and Silander 1975; Herrera 1987).

Results from floral bagging experiments evaluated one week after floral anthesis were inconclusive. There were some minor differences between several of the pollination treatments, but the pattern of incipient fruit set made no logical sense. Some flowers from all five of the treatments had a green, slightly swollen ovary, suggesting that they had been effectively pollinated and were developing into fruits. However, there are several alternative explanations. Because it is well known that jujubes have a low rate of fruit set, some growers spray a fertilizer (Komix®) at flowering time to enhance fruit set and

130

development (Nguyen, S.V., pers. comm., 201215). It is possible but unverified that this fertilizer treatment contributed to the ovary swelling observed in the experiment. It is also possible that ovary swelling may occur due to the presence of pollen. Neighbouring flowers receiving pollen may cause a cluster of flowers to behave as a resource sink, resulting in some ovary swelling. In addition, the slight ovary swelling that may result from pollination one week after anthesis may be impossible to record accurately, especially during Replication 1 when I had no prior experience. Finally, I may have accurately recorded those flowers that had been successfully pollinated, but only those flowers pollinated with appropriate pollen were retained by trees.

In contrast, by week 7 following anthesis, fruits remained only on branches receiving treatments that involved diurnal pollination (“open pollination” that allows for all modes of pollination and “diurnal pollination” with flowers bagged at night to allow only daytime visits by insects plus wind and self-pollination). The absence of fruits in treatments designed to detect wind pollination and self-pollination indicates that fruit set of jujubes (Z. mauritiana) must result from visits by daytime insect visitors between different jujube trees. In this study, that means that pollination was mostly a consequence of visits from honeybees and, to a much lower extent, flies.

In the first replication of the pollination experiment, it seemed initially that fruits would result from most of the pollination treatments. However, within a few weeks, all of those fruits had been aborted. This result is corroborated by some of the jujube growers who mentioned that early flowers usually set no fruit (Nguyen, H.C., pers. comm.,

15 Nguyen, S.V., Hanoi, Vietnam (+84 904 629 563).

131

201016). Later, low numbers of fruits were produced (Replication 2: 0.17%; Replication

3: 2.21%). One possible explanation for this differential fruit set is that there may have been insufficient pollination during the early part of the flowering period. During

Replication 1, due to the use of pesticides by growers, professional beekeepers dare not bring their hives to the site and the few honeybees kept in nearby villages were distributed among more than 40ha of jujube trees. Consequently, the major flower visitors during Replication 1 were flower flies and blow flies, and they may not have provided the necessary pollination services. By mid September when fruits were produced, the numbers of honeybees visiting jujube flowers had increased greatly. Yuan et al. (2009) suggested that insufficient pollination led to increased abortion of Ziziphus jujuba fruits. There may also be unknown physiological reasons underlying abortion of fruits early in the blooming period.

Jujube trees produce huge numbers of small flowers (Anonymous, n.d.) over an extended flowering period (Liu and Zhao 2008, Tel-Zur and Schneider 2009, Chapter 3).

It is normal that flowers and immature fruits are highly aborted in plants that produce an abundance of flowers (Stephenson 1981; Bawa and Webb 1984). Low resource availability and insufficient pollination (Stephenson 1981), limited pollen, incompatibility between pollen and stigma, selfing from the same clone or closely related plants (Piña et al. 2007), and high temperatures (Wani et al. 2010) are all possible reasons for fruit abortion. In Vietnam, jujube trees are usually grown from scions that originate from the same clone, yet jujube (Ziziphus mauritiana) is reported to be self-

16 Nguyen, H.C., Hanoi, Vietnam (+84 169 804 1955).

132

incompatible (Godara 1980b) and insufficient fertilization can reduce fruit yields (Yuan et al. 2009). The low fruit set observed in this study may be partially due to self- incompatibility issues. Unfortunately, the nature of the experiments reported in this thesis could not provide explanations for the patterns in fruit set and low numbers of fruits that matured.

Jujube growers receive 5000-6000 Vietnamese Dong per kg of fruits (0.24 – 0.29

USD per kg) (1.00 USD = 20,500 VND; Table 4.6). By combining fruit yields and their value, the value of the crop on a per tree basis could be estimated. On average for this site, the average yield across all cultivars was 92kg per tree with an average value of

5,700 VND per kg (0.28 USD per kg). With ~530 trees per ha, the average yield per ha was 48,760 kg, valued at 13,650 USD per ha. With A. cerana contributing ~97% of all pollination and the necessity of insect pollination of jujube flowers to set fruits, honeybees are estimated to have contributed a great proportion of this value.

Unfortunately, it was impossible to determine what fruit yields would have resulted from floral visits of flies and other insects in the absence of honeybees. Nevertheless, it is clear that honeybees substantially enhance the yields of jujube fruits and the profits to growers.

The main finding in this study is that only open pollination and diurnal pollination yield fruits and that jujube requires insects for pollination. Apis cerana is a critical pollinator that accounted for ~97% of pollination. Honeybees, and to a much lesser extent flies, performed close to 100% of all pollination of jujube flowers. Wind pollination, self- pollination, and nocturnal pollinators contributed nothing to jujube fruit set. Jujube

133

Table 4.6 Jujube yield and its economic value in Vietnam in 2011.

Mean yield Estimated Value Jujube per tree* contribution of Orchard cultivar (kg/tree) ± honeybees to gross VND/kg USD/tree SE income (USD)/tree**

F8 A 126 ± 12.9 6,000 36.9 35.9

F8 B 46 ± 1.8 6,000 13.5 13.1

F12 C 105 ± 6.4 5,000 25.5 24.8

*n = 8 trees per orchard **Values are 97.4% of values in the adjacent column.

growers should consider the role of honeybees in their production. Without honeybees, their fruit yields and consequently their profits would be greatly reduced. Reduced risk pesticides should be used sparingly when needed to enhance plant health or mitigate pest problems and as a result honeybees and other pollinators of jujube will be conserved.

Beekeepers should be welcomed to place their hives in or near jujube orchards.

134

CHAPTER 5. POLLINATION AND FRUIT SET OF LONGAN

(DIMOCARPUS LONGAN)

Abstract

The pollination biology of the tree fruit, longan (Dimocarpus longan), was studied in orchards in Hanoi, Vietnam, during two flowering seasons (2010 and 2011).

Longan inflorescences occur in thyrses with numerous flowers of three types: male (M1), hermaphrodite male (M2), and hermaphrodite female (F). Flowering in longan can be divided into three phases: male, female, and bisexual, with the order in which they occur varying between trees and cultivars. The diversity of floral visitors on longan flowers was relatively high. Of the total visits visually enumerated (n = 1851 in the first and n = 3364 in the second season), visits by honeybees were most frequent (47.4% and 14.0% of all visitors were A. cerana and A. mellifera, respectively, in the first season and 94.5% A. cerana only in the second season). Other floral insect visitors were primarily flies (both years) and moths (in 2010). Honeybees visited high thyrses more often than the low thyrses; other floral visitors visited low thyrses more than the high ones. Apis cerana had the highest index of pollinating efficacy (0.86 in 2010 and 0.99 in 2011). Bagging experiments revealed that longan set some fruits through wind pollination and self- pollination; however fruit set was greatest when diurnal visitors had access to flowers.

Apis cerana was a critical pollinator, augmenting longan production by ~67% in 2011.

The results suggest that insect pollinators, especially honeybees, need to be considered if growers are to obtain high yields of longan fruits.

135

5.1. Introduction

Several flowering plant species in the family Sapindaceae produce commercial fruits. Species in the family of high economic value include longan (Dimocarpus longan), litchi (Litchi chinensis), rambutan (Nephelium lappaceum), and pulasan (Nephelium mutabile) (Wong 2000; Nianhe and Gadex 2012). Within Sapindaceae, a mating system referred to as “duodichogamy” is widely distributed taxonomically. Within this system, individual trees are first functionally male (i.e., having only male flowers with functional pollen and rudimentary pistils), then functionally female (i.e., having only female flowers with fully developed pistils and short, non-functional anthers), and end their flowering with the production of hermaphrodite flowers that have well-developed pistils and stamens but that function as male flowers (Wong 2000; Davenport and Stern 2005;

Acevedo-Rodriguez et al. 2011). Self-incompatibility is common in the family, and separation of sexual phases within a tree prevents selfing (Acevedo-Rodriguez et al.

2011). Fragrant flowers with their abundant nectar attract many insects, including bees.

Pollen from the more common male and hermaphrodite flowers provide an additional reward for bee visitors (Acevedo-Rodriguez et al. 2011).

Longans are evergreen fruit trees grown throughout tropical Asia for their flavourful fruits (Crane and Walker 1984). In Vietnam, longans are grown widely in rural areas in both large commercial orchards and smaller home gardens (Fig. 5.1). The trees are visited by several species of insects, especially honeybees (Waite and Hwang 2002).

However, the contribution of flower visitors to pollination and fruit set of longans is poorly understood. Understanding of pollinators and floral biology is essential for longan growers to improve fruit yields.

136

Figure 5.1 A longan tree with numerous erect thyrses extending outward beyond the foliage.

Longans are important in Vietnamese culture and economy (Tran 2004, Sarfaraz

2011). Commercial plantings of longans have a total area of 98,000ha (Nguyen 2008) and are concentrated in several large areas: Hung Yen, Son La, Tuyen Quang, and Yen

Bai provinces in the North and Tien Giang province in the South (see map in Chapter 1,

Fig. 1.1). The value of longan fruits is estimated at 245 million USD annually. There are several cultivars of longans that differ in quality of the fruit (Morton and Miami 1987); these include “Chin Muon Hung Yen” (“Hung Yen”), “HTM1” (“Meo”), “HTM2

(“Tron”), and “Huong Chi” (Tran 2004; Tra 2012; Nguyen 2012). The size of orchards in different regions varies between hundreds of square meters to several hectares.

Floral buds of longans appear initially on very short shoots with several immature leaves at the ends of branches (Nakata et al. 2005, Davenport and Stern 2005). Longan flowers occur on 20-35cm long inflorescenses that are terminal, erect and without leaves.

137

The inflorescenses have three or four orders of monopodial axes, with dichasia consisting or 3, 7, or 15 flowers in each cluster. An inflorescence with this structure is referred to as a thyrse (see Fig. 2 in Nakata et al. 2005). Longan flowers are reported to be of three types: male flowers (M1), functionally female hermaphrodite flowers (F), and functionally male hermaphrodite flowers (M2). The female flower has two ovules.

Normally, only one ovule develops into a seed inside the fruit (Wong 2000, Davenport and Stern 2005).

Flowering in northern Vietnam usually occurs from March to April, but is sometimes delayed by about one month. During flowering, many kinds of insects visit flowers, including bees, flies, beetles, and moths (pers. observations). The nectar secreted by longan flowers probably serves as a reward for insect visitors. In turn, some of those insects pollinate the flowers. In a study from Queensland, Australia, both honeybees

(Apis mellifera) and stingless bees (Trigona spp.) increased initial longan fruit set

(Blanche et al. 2006b). Stingless bee visitation to longan flowers was more frequent when the longan trees were closer to forests, primarily due to the higher abundance of stingless bees in undisturbed habitats (Blanche et al. 2006b). Crane and Walker (1984) and Sarfaraz (2011) reported that longans may be self-incompatible, requiring cross- pollination. These studies suggest that insects play an important role in longan pollination. For example, honeybees (A. mellifera) have been suggested to boost longan fruit yield by as much as 30% (Crane and Walker 1984).

Longan flowers open at night, mostly from 0:00h to 6:00h. Anther dehiscence starts before noon, around 10:00h, at which time pollen becomes available to floral visitors. Stigma receptivity can be estimated by placing small droplets of hydrogen

138

peroxide on stigmas (Kearns and Inouye 1993). The appearance of small bubbles of oxygen and the rate at which they were produced indicated that stigmas were most receptive on the day of anthesis and the day after that: stigma receptivity lasts for several days, until the flowers senesce (Pham, H. D., unpublished data).

There is very little information on pollination of longans in the literature, and no information on the floral visitors and pollination of longans in Vietnam. Moreover, longans are a major edible fruit crop in Vietnam and an important source of nectar that honeybees convert into high quality honey. The objectives of this study were to determine: (a) the relative importance of types/modes of pollination that result in fertilization of longan flowers; (b) the relative importance of various potential pollinator species for longan yields, and (c) an estimate of the economic value of insect pollination on fruit yield.

5.2. Materials and Methods

5.2.1. Study sites and longan cultivars

The studies were carried out during two flowering seasons. In 2010, two longan orchards, one in Gia Lam district (25km East of Hanoi; 21.001N, 105.944E) and another in Quoc Oai district (30km West of Hanoi; 21.017N, 105.637E), were studied between 10

March and 30 April (Sunrise: 05:47h, Sunset: 18:12h, on 5 April). The vegetation in these districts consists of mixed crops including maize (Zea mays), rice (Oryza sativa), jujube (Ziziphus mauritiana), potato (Solanum tuberosum), mustard (Brassica hirta), pumpkin (Cucurbita spp.), eggplant (Solanum melongena), jackfruit (Artocarpus integrifolia), litchi (Litchi chinensis), Citrus spp., and ornamental plants. During the

139

experiments, the average daytime temperature recorded at the study site was 23.1oC

(17oC-29oC) and relative humidity (RH) was 71.2% (37%-84%) (personal data obtained between 06:00h-18:00h between 20 March and 12 April, 2010). Cultivars selected for the study in Gia Lam were “Chin Muon Ha Tay” and “Chin Muon Hung Yen”, which were referred to as “Ha Tay” and “Hung Yen”, respectively; those in Quoc Oai were “HTM1” and “HTM2”, which were referred to as “Meo” and “Tron”, respectively.

During 2010, numerous beekeepers moved their beehives to the study region to obtain longan honey. In Gia Lam, the number of beehives with A. cerana was estimated to be ~400; in Quoc Oai there were ~1200 hives of both A. cerana and A. mellifera bees in a village <1km from the orchard, along with ~100 hives with A. cerana in the longan study orchard and another 200 hives with A. mellifera ~50m away in an adjacent orchard.

In 2011, experiments were conducted in Quoc Oai district in two orchards, both of which contained both “Meo” and “Tron” cultivars. During the study period, the mean temperature was 24.7oC (21.5oC-32oC) and RH was 60.2% (29%-75%) (personal data obtained between 06:00h-18:00h between 11 April and 6 May, 2011). Weather ranged from sunny to cloudy, with light winds. There were several short rain showers, usually in the morning, on 8 of the 31 days of the study. In 2011, ~1000 A. cerana beehives were situated ~1km away in the nearby village.

5.2.2. Characteristics of flowers and flowering of longan

Types of flowers: Types of flowers were determined by visual observation. Floral morphology, colour, and numbers of stamens on male, female and hermaphrodite flowers were noted at various points in time on individual thyrses.

Floral dimensions: To quantify differences between flowers of different sexes,

140

five trees were randomly marked in one orchard in 2010. For eight male and eight female flowers on each tree, diameters of flowers, lengths of stamens, and length of styles were measured with callipers (hermaphrodite male flowers were not measured).

Thyrse structure: To fully understand the phenology of longan flowering, it was necessary to examine the structure of floral thyrses. They were described based on the branching pattern. The primary axis extended from the end of a tree branch itself; secondary axes branched off the primary axis; and tertiary axes were connected to secondary axes.

Sexual phases of flowering: Phases of flowering were noted by visual observation every 1-3 days, on 10, 12, and 8 trees of the “Ha Tay”, “Hung Yen”, and “Meo” cultivars, respectively, from start to end of flowering. Two thyrses per tree were studied to determine if flowering was in female phase (with only female flowers present), male phase (with only male flowers present), or in mixed phase with both male and female flowers. The duration of each sexual phase was also recorded.

5.2.3. Pollination treatments

In April and May 2010, an experiment on pollination of longans was carried out on four cultivars located in two orchards. In Gia Lam, five trees of “Ha Tay” and six trees from “Hung Yen” cultivars were randomly selected. On each tree, four thyrses of the same approximate size were tagged, and two or three spikes on each thyrse were marked. All female flowers on these spikes were studied. These flowers could be sexed as female because on the day before opening, the tips of the styles extended beyond the petals of the partially open flowers (Fig. 5.2). The four thyrses then received different bagging treatments (Blanche et al. 2006b): (1) exposure of flowers to all types of

141

pollination (open pollination; experimental control exposed continuously to insects, wind, and self pollination); (2) flowers enclosed by coarsely woven cheese cloth bags, with all anthers from marked flowers removed before they dehisced and released pollen, to quantify the effects of wind pollination; (3) flowers enclosed continuously in paper bag

(PBS International, Bag type: PBS 3d/60), allowing for self-pollination only; and (4) open pollination supplemented with pollination by hand. For hand pollination, trees that had male flowers were selected in orchards not included in other experiments. Newly opened male (M1 and M2) flowers were removed from these trees. Dehisced anthers on single male flowers were gently rubbed against the stigmas of individual female flowers that were receptive (i.e., in their first day after anthesis). Several stigmas were checked for presence of pollen grains on the stigma before and after this procedure to ensure that the hand pollination had been effective. Hand pollination was performed between 11:00h and 15:00h of days when newly opened female flowers were present on selected thyrses in the experimental trees. All bags and netting were applied on the day before anthesis.

Female flowers in the four treatments opened from 19 March to 30 March (mostly from

19 to 24 March). Data on fruit set were collected on 3, 8, 15, 29 April, 14 May, 13 June, and 13 July.

In Quoc Oai in 2010, five trees from “Meo” and four trees from “Tron” cultivars were randomly selected. The procedures were identical to those described above for the

Gia Lam orchard, except that a 5th pollination treatment was included: another thyrse on each tree was enclosed in mosquito-netting bags (mesh size excluded all insect visitors) at night, from 18:30h to 06:30h, but left exposed during the day to allow only diurnal pollination. All bags were applied on 19 March. Female flowers in the five treatments

142

opened from 20 to 30 March. Data on fruit set were also assessed on 7, 14, 28 April, 12

May, 12 June, and 12 July.

Figure 5.2 Morphology of female floral buds one day prior to anthesis. The floral buds with style partially extruded (indicated with an arrow) would open the next day.

In 2011, the study was conducted in Quoc Oai district from 10 April to 2 May, in two orchards with two cultivars, “Meo” and ”Tron”, in each orchard. Data were obtained from 7 “Meo” and 8 “Tron” cultivar trees in Orchard A, and 6 “Meo” and 11 “Tron” cultivar trees in Orchard B. On each tree, five thyrses of the same approximate size were tagged, and on each thyrse, short threads were tied around the pedicels of ~13 female flowers (range: 10-20) that would open the next day. The five thyrses then received different bagging treatments: (1) exposure of female flowers to all types of pollination

(open pollination; experimental control); (2) flowers (not emasculated) enclosed by coarsely woven cheese cloth bags, to quantify wind pollination; (3) flowers enclosed in mosquito-netting bags at night from 18:30h to 06:30h but open in the daytime (to allow

143

for diurnal pollination by insects as well as wind, and selfing); (4) flowers enclosed in mosquito-netting bags during the day, from 06:30h to 18:30h, but open at night (to allow for nocturnal pollination by insects, wind, and selfing); and (5) flowers enclosed continuously in paper pollination bags (PBS International, Bag type: PBS 3d/60) (to allow for self-pollination only). Because trees did not start the female phase simultaneously, bagging treatments were applied on the afternoon before the day of first floral anthesis. Data on fruit set were collected after 1, 2, 3, and 4 weeks, as well as on 18

June (~8 weeks), and 21 July (~12 weeks). For analyses, the data were standardized to the date of first anthesis.

5.2.4. Floral visitors

5.2.4.1. Determination of the community of flower-visiting insects

In 2010, four trees of each of the four cultivars were studied. Two thyrses at the same height on each tree were tagged. All insects observed on each floral branch were identified and counted by scan sampling (Dawkins 2007). The scans took 60min to complete and were initiated at 07:00h, 08:30h, 10:30h, 12:30h, 14:30h, 16:30h, and

18:00h in both orchards. The study was repeated on three days (01, 03, and 05 April). By the end of last scan of each day it was dark and observations were completed with the aid of headlamps. Representatives of floral visiting insects (except honeybees that could be identified in the field) were caught on 29 and 30 March and on 8 and 10 April; they were pinned for later identification.

In 2011, five trees of each of the two cultivars were randomly selected in both of the study orchards. To determine the effect of floral position on insect visitation, two

144

upper thyrses (height of ~3m) and two lower thyrses (height = ~1.5m) were selected and tagged on each tree. Scan sampling was employed again to quantify insect abundance and diversity. The scans took  90min to complete and were initiated at 07:00h, 09:00h,

11:00h, 13:00h, 15:00h, and 18:00h in both orchards. The study was repeated on six days

(26, 27, 28, and 30 April and 01 and 02 May). The last scans of the day were completed in the dark and observations were completed with the aid of headlamps. Representatives of floral visiting insects (except honeybees) were caught on 02, 03, and 04 May and pinned for later identification.

In both years, all insects were identified to family; in the case of honeybees and some of the larger flies, they were identified to genus or species by Dr. Stephen A.

Marshall. For the most common species, how they landed on flowers (e.g., their contact with various floral structures) and their movements on and between flowers were observed and recorded.

Durations of individual floral visits of the most common visitors were recorded between 06:00-09:00h, 09:00-12:00h, 12:00-17:00h, and 17:00-19:00h in 2011. The duration of a visit was defined by the elapsed time between when an insect landed on and left a flower. The number of visits per minute of a particular insect taxon was estimated by dividing 60 sec by the duration of an individual visit of that taxon. Travel time between flowers of the visitors was ignored in estimating this value.

5.2.4.2. Estimation of relative pollinating efficacy of main taxa of insect

floral visitors

Pollinator verification was based on specimens collected in 2011. To verify the presence of longan pollen on an insect’s body, each insect was placed in a separate, clean

145

Eppendorf tube (2.0ml), and bathed in 0.5ml or 1.0ml of 50% ethanol. The tube was vortexed three times for 12 sec each time. Two aliquots of 2µl each of the solution were withdrawn from the tube and loaded onto the left and right sides of a haemocytometer.

The haemocytometer was covered by a glass slide and then observed with a microscope at 400X. Pollen grains were counted in the four squares at the corners and one square at centre of the haemocytometer, then summed and divided by 5 to give the mean count per square. This procedure was replicated four times, with the tube containing the insect vortexed each time to ensure thorough mixing of pollen grains in solution. The haemocytometer was cleaned and dried after each count.

Total quantity of pollen on a single insect was calculated as follows:

Tp = (N × V1)/V2 where:

Tp: Total quantity of pollen grains

N: Average quantity of pollen grains on a large square of the haemocytometer

V1: Volume of dilution fluid (ml) (0.5ml or 1.0ml)

V2: Volume of a large square = 0.004µl

In the case of honeybees, pollen in their corbiculae was included in this measurement.

After removing pollen from the insects’ bodies, they were pinned for identification. Insect specimens were identified by Dr. Stephen A. Marshall, School of

Environmental Sciences, University of Guelph, Canada. Specimens are stored at the Bee

Research & Development Center in Hanoi, Vietnam.

Index of pollinating efficacy: An index of pollinating efficacy, I, was calculated for the four most common floral visitors by multiplying several different variables, as

146

follows:

I = V × P × L × F where:

I: Index of pollinating efficacy of a pollinator

V: Percentage of total floral visits by the specific pollinator species

P: Percentage of the insects of that species that had pollen on their bodies

L: Percentage of all pollen on the body that was longan pollen

F: Quantity of flowers that a pollinator visited per minute.

Values of I as calculated above assume equal contact by the different insect species with floral stigmas.

5.2.5. Influence of insect pollination on crop yield and economic efficacy

At harvest in the second season, 2011, fruits were collected from all selected trees

(as described on Section 5.2.3) within the two orchards. Yield (kg of fruits) per tree was recorded. The proportion of the fruit crop attributable to floral visits by honeybees was estimated as Phb = Ihb × (Pc − Pw − Ps) where:

Phb: proportion of the crop attributable to honeybees

Ihb : relative index of pollinating efficacy of honeybees

Pc: proportion of yield from diurnally pollinated flowers (treatment 3)

Pw: proportion of yield from wind pollinated flowers (treatment 2)

Ps: proportion of yield from flowers in bags that exclude insect visitation (self- pollinated flowers, treatment 5; equivalent to treatment 3 in 2010).

Pc, Pw, and Ps were calculated based on yield of fruits assessed on 21 July. The

147

proportions of Pc, Pw, and Ps sum to 1.

The portion of the gross income from sale of longans on a per tree basis attributable to honeybee pollination, Ghb, was estimated as:

Ghb = Phb × Yt × $/kg fruit where:

Yt : yield per tree.

5.2.6. Statistical analyses

To analyze floral visitors, data were modeled as a repeated measures generalized linear model using PROC GLIMMIX in SAS 9.2 (SAS Institute Inc., Cary, MA) with a

Poisson link function and autoregressive 1 (AR(1)) correlation structure. In 2010, data on floral visits in Quoc Oai was used. Random effects included day with repeated measures taken on trees nested within cultivar. Fixed effects included species of insect visitors (A. cerana, A. mellifera, and other visitors), treatments (“open” pollination and “open + hand” pollination), time periods (period 1: 7:00, 8:30, and 10:30; period 2: 12:30 and

14:30; period 3: 16:30 and 18:00), and cultivar. In 2011, random effects included day, orchard nested within cultivar and tree nested within orchard and cultivar. Fixed effects included species, treatment, observation time, and variety. All first order interaction terms between species, period or observation time and treatment were modeled. Pairwise comparisons between simple effects were adjusted for experiment-wise error rates using the Tukey-Kramer adjustment factor.

To analyze fruit set, data were modeled as a repeated measure generalized linear model using PROC GLIMMIX in SAS 9.2 (SAS Institute Inc., Cary, MA) with a Poisson link function and Toeplitz covariance structure. The repeated measures were taken on

148

uneven time intervals between days of data collection. The Toeplitz covariance structure was deemed to be the best fit. In the analysis of the data in 2010, random effects included tree nested within cultivar; fixed effects included treatment, time of observation, cultivar, and number of flowers (because the number of flowers varied, it had to be included for accurate interpretation of the numbers developing fruits). First order interaction effects included treatment by time of observation, treatment by cultivar and cultivar by time of observation. In the analysis of fruit set data in 2011, random effects included orchard nested within cultivar and tree nested within orchard and cultivar; fixed effects included treatment, time of observation and cultivar. First order interactions of cultivar were also modeled. Pairwise comparisons between simple effects were adjusted for experiment wise error rates using the Tukey-Kramer adjustment factor.

However, given the large number of comparisons, the significance of the difference for each comparison was interpreted at both 10% and 5% levels of significance.

5.3. Results

5.3.1. Characteristics of flowers and flowering of longan

Longan flowers consisted of three types, all occurring on the same thyrse: male

(M1; Fig. 5.3a), hermaphrodite male (M2; Fig. 5.3b), and hermaphrodite female that are functionally female (F; Fig 5.3c), as reported by Davenport and Stern (2005). The flowers were dull white to brownish yellow. M2 and F flowers had both male and female organs

(Fig. 5.3b and c).

M1 flowers measured 6.45 ± 0.636mm (n = 40) in diameter (petal tip to sepal tip).

149

Five sepals alternated with five petals; they surrounded a tiny pubescent protruding pistil at the center. The hairy stamens measured 4.23 ± 0.390mm (n = 40) in length. The number of stamens usually varied from six to nine (7.8 ± 0.08; n = 72) around the style.

Very rarely as many as 11 stamens were encountered.

M2 flowers were similar in appearance with M1 flowers except the pistil was

~20% longer than in M1 flowers and ended in a two-lobed light green stigma.

Female flowers (F) measured 7.07 ± 0.819mm (n = 40) in diameter and had an erect pistil with two ovules. The pubescent, bifurcated pistil was 2.96 ± 0.465mm in length (n = 40). Six to eight stamens surrounded the pistil. Those stamens were ~1 - 2mm in length, always shorter than the pistil, and never released pollen (Acevedo-Rodriguez et al. 2011).

a b c b Figure 5.3 Longan floral types. (A) The M1 flower with undeveloped pistil; (B) The M2 flower has an easily seen pistil that is smaller than the pistil of the F flower; (C) The F flower has fully developed pistil with two lobed stigma. Stamens in F flowers were ~1-2 mm in length and always shorter than the pistil.

Longan inflorescences are complex and occur in what are referred to as thyrses

(Nakata et al. 2005). Each thyrse has a primary axis that is an extension of a tree branch.

Secondary axes, sometimes referred to as spikes, are attached alternately to the primary axis (Fig. 5.4, left). Similarly, tertiary axes attach alternately on the secondary axes. The

150

tertiary axes consist of clusters of flowers referred to as dichasia. At the terminal end of each dichasium is a primary flower. Basal to each primary flower and on either side are two branches, with terminal secondary flowers and two tertiary flowers opposite each other in more basal positions (Fig. 5.4, right, numbering 1, 2, and 3 respectively). The primary flowers on a thyrse could be either M1 or F flowers (see below) (Fig. 5.5). Some variation between trees in numbers of tertiary and quarternary flowers within dichasia was observed but these differences were not systematically recorded.

Figure 5.4 A). Diagram of longan thyrse showing lateral spikes and floral buds. The letters show primary (a), secondary (b), and tertiary (c) axes. A secondary axis is equivalent to a spike. B). Detail of tertiary axis, or dichasium with numbers showing the sequence of floral anthesis. The relative sizes of the buds reflect their approximate sizes on the day of anthesis of the primary flower (redrawn from Nakata et al., 2005).

Flowering of longan trees could be divided into three phases: male (M1 and/or

M2 flowers), female (F flowers only), and both sexes (M and F flowers), depending on

151

sex of flowers on a tree at any given time. The sequence of phases varied among trees and cultivars. A majority of “Hung Yen”, “Ha Tay”, and “Meo” cultivar trees (22/30 =

73.3%) began with all primary flowers opening in female phase. The female phase usually lasted for about three to five days, following which they changed to male phase until the end of the flowering period. However, a few additional trees (2/30 = 6.7%) returned to the female phase again for the last few days at the end of the flowering period.

Usually there was an interruption in flowering for a few days within thyrses as they shifted their flowering phase. Some trees (5/30 = 16.7%) started with both male and female phases concurrently, although the primary flowers within a single thyrse were all of the same sex. Occasionally (1/30 trees = 3.3%) all flowers on a tree were male and no fruits were produced. In contrast, “Tron” cultivar trees began flowering with male primary flowers; this male phase lasted from a few days to two weeks before they shifted to female phase. The female phase lasted for 4 to 5 days, then they returned to male phase near the end of the flowering period. In this part of my research, M1 and M2 flowers were not differentiated because they were both functionally male.

Figure 5.5 Primary flowers opening in female phase (left) and male phase (right).

152

5.3.2. Effect of insect visitation on fruit set

Pollination experiments were carried out in both years, with treatments depending on year and location. In 2010, all treatments in the experiment in Gia Lam yielded some fruits, except self-pollinated flowers (the last fruit from the self-pollination treatment was aborted between 29 April and 14 May) (Fig. 5.6). Most incipient fruits were aborted within the first three assessments, from 3 to 15 April, especially those on inflorescences receiving the wind pollination and self-pollination treatments. There was an interaction between treatments and times of observation (F = 2.20; df = 18, 250; P = 0.0040). Fruit set evaluated after just 1 week (on 3 April) did not differ between treatments (all P values

> 0.05). By 8, 15, and 29 April, open pollination and open + hand pollination treatments had more fruits than wind pollination and self-pollination treatments (all unadjusted P values < 0.05). Fruit set from open and open + hand pollination treatments was constantly greater than that from wind pollination treatment to the last assessment 13 July (all unadjusted P values < 0.05). Open pollination and open + hand pollination treatments never differed in fruit set at any time during the study (all P values > 0.05). Four flowers from the open + hand pollination treatment had two fruits per flower on 13 July.

There was also a significant interaction between cultivar and date of observation in the number of developing fruits (Fig. 5.7) (F = 3.42; df = 6, 250; P = 0.003). Number of fruits between “Ha Tay” and “Hung Yen” cultivar did not differ on 3 April, but assessments on 8 and 15 April, 13 June, and 13 July indicated that “Ha Tay” cultivar had more fruits (all unadjusted P values < 0.05). However, within cultivar and between all

153

40 a a a Wind pollination Open & hand pollination

35 Open pollination Self-pollination 30

25 a

a 20 a a 15 a a a a a a a a 10 a b 5 b b b b b b b b

Mean number of developing fruits per thyrseper fruits developing of Meannumber 0 3 April 8 April 15 April 29 April 14 May 13 June 13 July

Observation dates

Figure 5.6 Mean number of developing fruits recorded over time in Gia Lam, 2010. All treatments yielded some fruits, except self-pollinated flowers (the last fruit from the self- pollination treatment was aborted between 29 April and 14 May). By 8, 15, and 29 April, open pollination and open + hand pollination treatments had more fruits than wind pollination and self-pollination treatments. Fruit set from open and open + hand pollination treatments was continuously greater than that from wind pollination treatment to the last assessment 13 July (all unadjusted P values < 0.05). Data are the mean numbers of developing fruits per thyrse, with 95% confidence intervals (n = 9 for wind pollination treatment, n = 11 for open, open and hand, and self-pollination treatments at each time of observation). Within an observation date, the same letters indicate no significant difference.

successive time intervals, number of incipient fruits differed only between 3 and 8 April and between 8 and 15 April (all unadjusted P values < 0.05). This result indicates that the number of incipient fruits decreased rapidly over the first three weeks after the experiment was established, after which fruit abortion was much less (Fig. 5.7).

154

35 a Ha Tay Hung Yen a 30

20 a thyrse 15 a a 10 b a a 5 b a

Mean number of developing fruits per fruits developing of Meannumber a b a b 0 3 April 8 April 15 April 29 April 14 May 13 June 13 July Observation dates

Figure 5.7 Mean number of developing fruits recorded on different dates from the “Ha Tay” and the “Hung Yen” cultivars in Gia Lam, 2010. Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 18 for the “Ha Tay” cultivar and n = 24 for the “Hung Yen” cultivar). Lower and upper confidence intervals for the “Ha Tay” cultivar on 13 June and 13 July were (0.002431, 45.4253) and (0.000283, 278.16), respectively and lower and upper confidence intervals for the “Hung Yen” cultivar on 13 June and 13 July were (0.001349, 25.0298) and (0.000157, 153.67), respectively. Within an observation date, the same letters indicate no significant difference.

In Quoc Oai in 2010, flowers within all treatments yielded some fruits at each time of observation (Fig. 5.8). Many incipient fruits dropped off between 14 to 28 April, especially from the self-pollination treatment. There were significant differences in fruit set between dates of observation (F = 20.87; df = 5, 255; P < 0.0001) and cultivar (F =

13.86; df = 1, 7; P = 0.007). However, due to insufficient degrees of freedom to analyze interactions between treatments and times, the data had to be analyzed separately for each observation date. There were no significant differences in incipient fruits from the five treatments on 7 April (the first assessment) (F = 0.44; df = 4, 31; P = 0.775). By 14 April, fruit set in the open pollination treatment differed from the self-pollination treatment

155

(P = 0.022). From the third assessment on 28 April through to the last assessment on 12

July, fruit set in the open pollination and diurnal pollination treatments exceeded fruit set in the self-pollination treatment (P < 0.05). In the fourth assessment on 12 May, fruit set from open and hand pollination treatment also differed with that from self-pollination

(P = 0.038). No significant differences in fruit set were detected between open pollination and open + hand pollination treatments at any time in this study (P > 0.05). From Figure

5.8, it is evident that many fruits dropped between 14 and 28 April, and more fruits were aborted in the self-pollination treatment than in the other treatments. On 12 July, 2 fruits derived from a double-fertilized flower in the self-pollination treatment were still present.

Diurnal pollination Wind pollination Open & hand pollination Open pollination Self-pollination 10 a 9 ab ab 8 7 a a a a 6 ab ab a a a 5 ab b ab a a 4 ab ab ab ab 3 b 2 b b b 1 0 Mean numbers of developing fruits per thyrseper fruits developing of Meannumbers 14 April 28 April 12 May 12 June 12 July Observation dates

Figure 5.8 Developing fruits recorded at different times in Quoc Oai in 2010. All treatments yielded some fruits at each time of observation. Fruit set from open pollination and diurnal pollination treatments differed with that from self-pollination from the third assessment on 28 April through to the last assessment on 12 July (P < 0.05). Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 9 for each treatment at each time of observation). Within an observation date, the same letters indicate no significant difference.

156

In 2011, all treatments had some fruits up to the last check on 21 July (week 12).

As in the preceding year, incipient fruits were aborted mostly in the first three weeks

(Fig. 5.9). Data were analyzed only until 21 July because after that rats destroyed many fruits as they matured. There was an interaction between treatments and dates of observation (F = 2.76; df = 20, 876; P < 0.0001). In the first assessment, significant differences in incipient fruit set were found between the open pollination treatment and nocturnal pollination (P = 0.001), wind pollination (P < 0.0001), and self-pollination treatments (P < 0.0001) and between diurnal pollination treatment and wind pollination

(P = 0.0008) and self-pollination treatments (P = 0.001). The differences between treatments changed over time. At week 2, the number of incipient fruits differed between diurnal pollination and wind pollination treatments (P = 0.034), between diurnal pollination and self-pollination treatments (P < 0.0001), and between open pollination and self-pollination treatments (P < 0.0001). By week 3, significant differences were detected between diurnal pollination and self-pollination treatments (P = 0.002) and between open pollination and self-pollination treatments (P = 0.005). At week 4, fruit set from diurnal pollination and open pollination treatments differed with nocturnal pollination, wind pollination, and self-pollination treatments (all P values < 0.05). On 18

June (week 8), diurnal pollination differed from three other treatments: nocturnal pollination, wind pollination, and self-pollination (all P values < 0.05). However, in the last assessment on 21 July (week 12), due to reduced numbers of fruits in all treatments, only two treatments still differed: diurnal pollination and self-pollination (P = 0.041). In summary, diurnal pollination and/or open pollination had more fruits than one or more of the three other pollination treatments (nocturnal pollination, wind pollination, and self- pollination) during at least one of the assessments.

157

8 a Nocturnal pollination Diurnal pollination

Wind pollination Open pollination

7 ab Self-pollination

5 bc

4 c c

3 a ab 2 a a a a abc bc a a ab ab 1 ab ab c b b b b ab b b ab ab b Mean number of developing fruits per thyrseper fruits developing of Meannumber 0 Week 1 Week 2 Week 3 Week 4 Week 8 Week 12 Week of observation

Figure 5.9 Developing fruits recorded at different times in Quoc Oai in 2011. All treatments yielded some fruits up to week 12. Incipient fruits were aborted mostly in the first three weeks. Diurnal pollination and/or open pollination had more fruits than one or more of the three other pollination treatments (nocturnal pollination, wind pollination, and self-pollination) during at least one of the assessments. Data are the mean numbers of developing fruits, with 95% confidence intervals (n = 31 for nocturnal pollination and wind pollination treatments; n = 32 for open pollination and self-pollination treatments at each time of observation; in diurnal pollination treatment: n = 32 for the first four weeks and n = 31 for week 8 (18 June) and week 12 (21 July). Within an observation week, the same letters indicate no significant difference.

In 2011, there were significant interactions in fruit set between cultivars and dates of observation (F = 6.77; df = 5, 876; P < 0.0001) (Fig. 5.10). In comparisons between all successive time intervals, within the “Meo” cultivar, numbers of incipient fruits declined between week1 and week 2, between week 2 and week 3, and between week 3 and week

4 (all P values < 0.05). Similarly, the “Tron” cultivar also had significant changes in fruit set between week 1 and week 2 and between week 2 and week 3 (both P values < 0.05).

158

The “Meo” cultivar had more incipient fruits than the “Tron” cultivar at week 1, shortly after flowering period (Fig. 5.10, P = 0.024). That difference became insignificant by week 2 (P = 0.512).

12 Meo Tron a 10

4 b a 2 a a a a a a a a a 0 Mean number of developing fruits per thyrseper fruits developing of Meannumber Week 1 Week 2 Week 3 Week 4 Week 8 Week 12

Week of observation

Figure 5.10 Developing fruits recorded at different time periods for the “Meo” and the “Tron” cultivars, 2011. Data are the mean numbers of developing fruits, with 95% confidence intervals (in “Meo” cultivar, n = 13 trees; in “Tron” cultivar, n = 19 trees). Within an observation week, the same letters indicate no significant difference. The last two checks were performed on 18 June and 21 July.

5.3.3. Diversity, abundance and pollination efficacy of floral visitors

5.3.3.1. Diversity of floral visitors

Data on the diversity and proportions of insect taxa visiting flowers were obtained by both scan sampling and by capturing specimens for identification. Specimens of 41 families in 8 orders of insects were caught while visiting longan flowers at the study sites in the two seasons (Table 5.1). Most belonged to the orders Hymenoptera, Diptera and

159

Lepidoptera. The greatest diversity, with 13 families represented, were flies; followed by

Hymenoptera, Hemiptera, and Coleoptera with 7, 7, and 6 families, respectively. The most common visitors were honeybees (A. cerana and A. mellifera), some taxa of flies

(blow flies, house flies, and flower flies), and several families of moths in the early evening. Most others were rarely observed. Of the total visits visually enumerated during scan sampling (n = 1851 scans in the first year and n = 3364 in the second year), visits by honeybees were most frequent. Apis cerana and A. mellifera accounted for 47.4% and

14%, respectively, in 2010, and 94.5% (A. cerana only) in 2011. The remainder was mostly calliphorids, syrphids, muscids, and in 2010 several families of moths.

Table 5.1 Insect visitors captured on longan flowers.

Number of specimens* Order Family/subfamily Genera and Species 2010 2011 Blattodea Blatellidae 15 (2) 4 Coleoptera Bruchidae 2 0 Coleoptera Chrysomelidae 2 0 Coleoptera Coccinellidae 1 0 Coleoptera Elateridae 1 0 Coleoptera Latridiidae 6 0 Coleoptera Scarabaeidae 1 (2) 1 Diptera Anthomyidae 1 0 Calliphoridae: Calliphorinae, Chrysomyia Diptera 32 (291) 30 (60) Crysomyinae, and megacephala Lucilliinae Diptera Chironomidae 40 (6) (5) Diptera Culicidae 3 0 Diptera Muscidae Musca spp. 17 (115) 34 (98) Diptera Platystomatidae 1 0 Diptera Rhiniidae 25 0 Diptera Sarcophagidae 2 0

160

Number of specimens* Order Family/subfamily Genera and Species 2010 2011 Diptera Sciomyzidae Sepedon sp. 1 0 Diptera Sepsidae 1 0 Diptera Stratiomyidae 4 0 Eristalinus lathyrophtalinus, Syrphidae: E. mesembrius, Diptera Syrphinae, 29 (93) 3 (6) E. megacephalus, Eristalinae Episyrphus balteatus Diptera Tipulidae 1 0 Hemiptera Anthocoridae 4 0 Hemiptera Aphididae 1 0 Hemiptera Cicadellidae 1 0 Hemiptera Lygaeidae 1 0 Hemiptera Miridae 11 0 Hemiptera Pentatomidae 1 0 Hemiptera Pyrrhocoridae 2 (1) A. cerana, (877) (3178) Hymenoptera Apidae A. mellifera, (260) (0) A. florae (4) (0) Hymenoptera Braconidae 4 0 Hymenoptera Chalcididae 4 0 Hymenoptera Crabronidae 2 (47) 0 Hymenoptera Ichneumonidae 3 0 Hymenoptera Scoliidae 2 0 Vespidae: Hymenoptera 3 (2) (1) Eumeninae Lepidoptera Arctiidae 1 Lepidoptera Geometridae 2 Lepidoptera Noctuidae 2 (145) (15) Lepidoptera Pieridae 1 Lepidoptera Pyralidae 7 Neuroptera Chrysopidae 4 (3) 0 Thysanoptera Thripidae (4) 0 *Numbers in parentheses were number of records obtained from scans.

161

5.3.3.2. Floral visitor behaviour

Visits of floral visitors varied between times of day and between years (Fig.

5.11a, b). Honeybee (Apis spp.) visits differed during the day. In 2010, visits in the morning (7:00h- 11:30h) were low, but they increased shortly after noon (12:30h-15:30h) and declined in the late afternoon (A. cerana, 16:30h-19:00h). There was an interaction in numbers of visits between A. cerana, A. mellifera, and other visitors and different times of days (F = 2.97; df = 4, 855; P < 0.0187). Fig. 5.11a shows that the temporal patterns exhibited by Apis cerana and A. mellifera were similar, in contrast to the pattern of visits by other insect visitors that increased from morning to afternoon and did not differ between afternoon (predominantly flies) and evening (predominantly moths) (P > 0.05).

In 2011, the numbers and pattern of visits by honeybees and other visitors (Fig.

5.11b) were strikingly different from those recorded in 2010. Apis cerana dominated the community of flower visitors while the numbers of visits by other insects was insubstantial. There was a highly significant interaction between numbers of visits per thyrse and time of day (F = 21.28; df = 5, 5715; P < 0.0001). Visits by honeybees differed between each pair of time intervals (P < 0.05), except for the periods beginning at 09:00 and 11:00h (P = 0.599). Their activity peaked earlier than in 2010, between

9:00h and 11:00h. In contrast, diurnal visits by the other main taxa of insect visitors (data pooled) were much fewer in number than in 2010 and did not vary temporally (P > 0.05 for all pairwise comparisons).

162

4 Apis cerana a 3.5 Apis mellifera 3 Other insect visitors abc b ab 2.5

2 cd 1.5 de 1 def def ef 0.5

0 Mean number of insect visitors per thyrse per insect ofvisitors Meannumber 7:00-11:30h 12:30-15:30h 16:30-19:00h

a) Observation periods

Apis cerana Other insect visitors

4.5 a 4 a 3.5 3 b 2.5 c 2 1.5 d 1 0.5 e f f f f ef ef 0 Mean number of insect visitors per thyrse per insect ofvisitors Meannumber 07:00h 09:00h 11:00h 13:00h 15:00h 18:00h

b) Start times of observation period

Figure 5.11 Mean numbers of Apis cerana and other insect visitors per thyrse during 2010 (a) and 2011 (b) flowering seasons. The pattern of diurnal activity and abundance of honeybees in the two seasons differed. Visits by honeybees peaked in the afternoon in 2010 (a) compared to the morning in 2011 (b). Numbers of other visitors observed on longan flowers in 2011 were much fewer than that in 2010. Data points are the mean numbers of visitors with 95% confidence intervals observed during 286, 191, and 191 scans in each time interval in 2010 and 480 scans in each time interval in the 2011 season. The same letters indicate no significant difference.

163

Hand pollination of flowers had no effects on insect visitation. Flowers in the open pollination treatment were visited at the same frequeny as those that received supplemental hand pollination (F = 2.04; df = 1, 855; P = 0.154) (Fig. 5.12).

1.4 a

1.2 a 1

0.8

0.6

0.4

0.2

Mean number of insect visitors per thyrse per insect ofvisitors Meannumber 0 Open Open + hand Treatments

Figure 5.12 Mean numbers of insect visitors to thyrses receiving pollination treatments in 2010 observed during scan sampling, with 95% confidence intervals (n = 166 scans of flowering thyrses in each of the open and open + hand pollination treatments). The same letters indicate no significant difference.

In 2011, floral visitors differed in their visitation to upper and lower thyrses on the trees. Specifically, there was a significant interaction between numbers of visits of honeybees and other species and the positions of thyrses on trees (F = 93.17, df = 1,

5715; P < 0.0001). Higher thyrses were visited by A. cerana more often than lower thyrses (P < 0.0001). In contrast, other species of insects visited lower thyrses in greater numbers than the higher ones (P < 0.0001) (Fig. 5.13).

164

2.5 a Apis cerana

2 Other insect visitors

1.5 b 1

0.5 c

Mean number of visitors per visitors ofthyrse Meannumber d 0 High Low

Positions of thyrses

Figure 5.13 Number of Apis cerana and other insect visitors recorded on longan thyrses differing in location on the trees in 2011. Floral visitors differed in their visitation to upper and lower thyrses on the trees. Higher thyrses were visited by A. cerana more often than lower thyrses. Other species of insects visited lower thyrses in greater numbers than the higher ones (P < 0.0001). Data points are the mean numbers of visitors observed during scan sampling, with 95% confidence intervals (n = 720 scans of flowering thyrses in each position of thyrse for two groups of insect visitors). The same letters indicate no significant difference.

In the early morning when the temperature was relatively low and RH was relatively high (22.4oC and 72% at 06:00h), there was relatively little activity on flowers compared to later observation periods (Fig. 5.14). A weak interaction between times of day and cultivar was detected (F = 2.39; df = 5, 5715; P = 0.036). Number of visits between each pair of time intervals and within each cultivar did not differ (all P values >

0.05), except for the period beginning at 7:00h (P < 0.0001). Also, within each time interval and between the two cultivars, number of visits did not differ (all P values >

0.05).

165

0.7 Meo Tron 0.6 a a a a a 0.5 a a a 0.4 a a

0.3

0.2 b

b Number of visits per panicle per Numbervisits of 0.1

0 07:00h 09:00h 11:00h 13:00h 15:00h 18:00h Starting times of observation period

The four main insect taxa that visited longan flowers in 2011 differed in the duration of their visits over the course of the day (F = 1.99; df = 9, 848; P = 0.037; Fig.

5.15). Average durations of floral visits for A. cerana bees, flower flies (Syrphidae), and blow flies (Calliphoridae) were 5.15 ± 5.19 sec, 8.31 ± 4.56 sec, and 8.90 ± 4.13 sec, respectively, and did not differ significantly. Houseflies remained on flowers for significantly longer periods of time (33.13 ± 5.31 sec). Of the four main taxa studied, the duration of visits by A. cerana, flower flies, and house flies did not vary during the day, but blow flies stayed on flowers significantly longer in mid morning than at other times of day.

166

A. cerana Blow flies Flower flies House flies

45 60a 60a 40 51ab 60ab 35 30 25 20 59abc 15 52d 17abcd 48bcd 10 54d 36bcd 48cd 60d 5 77cd 61cd 40cd 91cd Mean duration of floral visits visits (sec)floralof Meanduration 0 -5 06:00-09:00h 09:00-12:00h 13:00-17:00h 17:00-19:00h Observation intervals

Figure 5.15 Mean (± SE) duration (sec) of floral visits of four taxa of insect visitors to longan flowers at different times of day. Means with the same letters are not significantly different (P > 0.05, Tukey’s test). Duration of visits for A. cerana, flower flies (Syrphidae), and blow flies (Calliphoridae) did not vary during the day. Houseflies remained on flowers for significantly longer periods of time than the three other taxa.

Generally, all of the primary insect visitors tended to move between flowers within a thyrse. When they landed on flowers, they clung to floral parts (sepals, petals, stamens, and pistil) while probing for nectar. Their forelegs and/or mid-legs touched dehisced anthers and pollen became stuck to their legs or other parts of their bodies.

When they foraged on an adjacent flower, their legs and proboscis frequently came in contact with the stigmatic surface and pollen was presumably deposited there. Apis cerana always turned around while probing flowers for nectar. They contacted anthers and/or stigmas of most flowers (18/19 = 94.7%) while they foraged. The three other taxa contacted floral reproductive organs less frequently (61.5%; n = 39).

Moths that visited flowers in moderately high numbers in the early evening during

2010 rarely touched anthers or stigmas while foraging. Some hovered in front of flowers

167

while others rested on a floral branch as they probed nearby flowers for nectar with their proboscis (Fig. 5.16). Coleopterans moved slowly among flowers. Cockroaches

(Blatellidae) usually emerged after dark and walked along branches. When they visited flowers, they spent several minutes collecting nectar from the sides of flowers without touching stigmas to any appreciable extent (Fig. 5.16).

Figure 5.16 Floral visitors of longan: Top left: A honeybee, Apis cerana; top right: a flower fly; middle left: a blow fly; middle right: a house fly; bottom left: a cockroach; and bottom right: a moth at rest on a floral branch with proboscis extended.

168

5.3.3.3. Pollen loads of visitors

The primary longan visitors were divided into four groups: honeybees (A.

cerana), Syrphidae (flower flies), Calliphoridae (blow flies), and Muscidae (house flies

and their relatives). Among these floral visitors, only honeybees had large numbers of

pollen grains on their bodies (Table 5.2). The flies in the three other taxa had relatively

small amounts of pollen on them. A high percentage of the honeybees and the flower flies

had pollen on their bodies (78.6% and 75.0%, respectively), whereas less than half of the

other flies did. However, if they had pollen on them, it was 95-100% longan pollen,

reflecting a high degree of floral constancy while foraging or a lack of alternate floral

resources.

Table 5.2 Percentage of insects that carried pollen and quantity of longan pollen on specimens sampled on longan flowers.

No. of insects that Mean No. of Mean No. of carried pollen Percentage of Foragers N pollen grains longan pollen longan pollen (± SE) grains (± SE) N %

Honeybee (A. 220,000 ± 220,000 ± 56 44 78.6 100 cerana, Apidae) 35,000 35,000 Flower flies 8 6 75.0 6,300 ± 2,000 6,300 ± 2,000 100 (Syrphidae) Blow flies 29 11 37.9 5,000 ± 1,600 5,000 ± 1,600 100 (Calliphoridae) House flies 33 10 30.3 4,100 ± 1,200 3,900 ± 1,200 95.1 (Muscidae)

169

5.3.3.4. Index of pollinating efficacy

The relative pollinating efficacy of the four main taxa of pollinators was estimated, based on the relative abundance of each type of insect on longan flowers (data obtained from scan sampling of thyrses in 2010 and 2011), the amount of longan pollen that the visitors carried, and the duration of floral visits. Honeybees (A. cerana) were by far the most important pollinator, estimated to have contributed 86% in 2010 and 95% in

2011 of all insect pollination. They were followed in importance by blow flies, flower flies (Eristalinus megacephalus and E. lathyrophthalinus), and house flies (Table 5.3).

Indices of pollination efficacy were not estimated for A. mellifera and moths in 2010 because no data on pollen on their bodies were obtained. Indices of pollination efficacy were not calculated for other insect visitors due to their scarcity. Estimates of relative pollinating efficacy for floral visitors in 2010 were made using the data obtained in 2011 on the proportion of insects with pollen, the proportion of longan pollen, and the numbers of flowers visited per min, with an assumption that these data were consistent between years.

5.3.4. Economic value of insect pollination

Yields of fruits from different pollination treatments were calculated based on data on fruit set assessed on 21 July, 2011. The yields from the diurnal pollination, wind pollination, and self-pollination treatments were 0.248, 0.025, and 0.023 per thyrse, respectively (Fig. 5.9, 21 July); the relative proportions of these three treatment yields were 0.838, 0.084, and 0.077, respectively. Fruit yield data were not obtained in 2010.

170

Table 5.3 Pollinating efficacy of Apis cerana and other flower visitors on longan flowers.

Flowers Proportion Proportion Relative visited Proportion Pollination Pollinators with of longan Year pollinating per of visitors index* pollen pollen efficacy** minute

Honeybee 2010 0.47 4.34 0.860 (A. cerana, 0.79 1.0 11.7 2011 0.95 8.78 0.991 Apidae)

Blow flies 2010 0.16 0.41 0.081 0.38 1.0 6.7 (Calliphoridae) 2011 0.02 0.05 0.006

Flower flies 2010 0.05 0.27 0.053 0.75 1.0 7.2 (Syrphidae) 2011 0.002 0.01 0.001

House flies 2010 0.06 0.03 0.006 0.30 0.95 1.8 (Muscidae) 2011 0.03 0.02 0.002

*Pollination index is the product of the four variables; the proportion of visitors was derived from scan sampling data. **The relative values, with relative pollination values for the four main taxa of pollinators; within year, these values sum to 1.

Honeybees are only active in the daytime; therefore, the yields attributed to

honeybee pollination were calculated using the relative value for diurnal pollination ,

minus the contributions from wind and self-pollination). Consequently, the proportion of

the longan yields attributable to floral visits by honeybees (Phb) was calculated as

Phb = Ihb × (Pc − Pw − Ps)

where Ihb is index of pollination efficacy of honeybees in 2011, Pc is proportion of the

fruit yield from diurnally exposed flowers (treatment 3), Pw is proportion of fruit yield

171

from wind pollinated flowers (treatment 2), and Ps is proportion of yield from flowers in bags that excluded insect visitation (self-pollinated flowers, treatment 5).

Phb = Ihb × (Pc − Pw − Ps)

= 0.99 × (0.839 − 0.084 − 0.077) = 0.99 × 0.678

= 0.67

A similar value for 2010, if it could be computed, would be smaller because the index of pollination efficacy for honeybees in 2010 was lower than in 2011.

Harvest of longans started in late August and finished by mid-October, 2011. The average yield for the “Meo” cultivar trees was 124 ± 10.2kg per tree in orchard A and 74

± 9.0kg per tree in orchard B. The average yield for the “Tron” cultivar trees was 120 ±

10.6kg per tree in orchard A and 80 ± 6.1kg per tree in orchard B (Table 5.4).

Table 5.4 Longan yield in Quoc Oai 2011 and its economic value.

Yield per Estimated contribution Value (USD) tree* of honeybees to gross Cultivar Orchard (kg/tree) ± income (USD)/tree** Per kg Per tree SE

“Meo” A 124 ± 10.2 0.73 90.5 60.6

“Meo” B 74 ± 9.0 0.73 54.0 36.2

“Tron” A 120 ± 10.6 0.73 87.6 58.7

“Tron” B 80 ± 6.1 0.73 58.4 39.1

*n = 8 trees per orchard **Values are 67% of values in the adjacent column and were calculated from the values of pollination treatments in the fruit set experiment assessed on 21 July 2011).

Longan growers receive 15,000 Vietnamese Dong per kg for fruits (0.73 USD per kg) (1.00 USD = 20,500 VND; Table 5.4). By combining fruit yields and their value, the

172

crop on a per tree basis could be estimated. On average for this site, the average yield across all cultivars was 100kg per tree. With honeybees contributing approximately 67% of longan pollination at this location, honeybees are estimated to have contributed pollination services worth ~48.70 USD per tree in 2011. Unfortunately, it was impossible to determine what fruit yields would have been as a result of floral visits of flies and other insects in the absence of honeybees. However, it is clear that honeybees substantially enhance the yields of longan fruits and profits to growers.

5.4. Discussion

Bagging experiments indicated that all pollination treatments yielded some mature fruits, regardless of location and year of study, except in the self-pollination treatment in Gia Lam, 2010. However, numbers of fruits maturing from open, open + hand, and diurnal pollination treatments were greater than numbers obtained from nocturnal, wind, and self-pollination treatments, during at least one assessment.

Supplemental hand pollination did not increase fruit set in comparison with continuously exposed flowers (i.e., those in the open pollination treatment) (see below). Visits by diurnal pollinators were very important in obtaining high yields of longans, also detailed below.

Numbers of developing fruits varied by time of assessments. When flowers were first checked two weeks after floral anthesis, fruit set results were inconclusive. At this time there were only minor differences between treatments in numbers of flowers with green, slightly swollen ovaries that appeared to have been effectively pollinated and were developing into fruits. However, fruit abortion in the first few weeks after flowering was extensive and continued until fruit maturity. Others have also reported that physiological

173

fruit abortion occurs in longans, particularly in the first few weeks after flowering

(MeiYing et al. 2004, WenBin 1999). There are several possible explanations for fruit abortion, including resource limitation, limited pollen arriving on stigmas, and extensive self-pollination (fruits from self-pollinated flowers tend to abort more often than fruits developing from cross-pollinated flowers; Stephenson 1981). Similar levels of fruit abortion occur in litchi (Mitra et al. 2005; JinHua 2007). By the last assessments of fruit set, very few fruits remained on trees except on thyrses that received the open treatment, open and hand pollination treatment (in 2010), and diurnal pollination treatment (in

2011).

In 2010, insect visitation and fruit set were compared between flowers continuously exposed to all types of pollination (open treatment) and exposed flowers supplemented with hand pollination (open + hand treatment). Insect visitors were quantified on the flowers in these two treaments through scan sampling ~2 days after being hand pollinated. They received equal numbers of visitors as the flowers in the open pollination treatment. Importantly, there was also no difference in fruit set between these two treatments. That there was no improvement of fruit set with hand pollination suggests that pollination in the study orchards was not a limiting factor for fruit production. In other words, flowers in the open pollination treatment (exposed to diurnal and nocturnal pollinators, wind pollination, and self-pollination) matured fruits at a level at the physiological limit of the longan trees.

The most important result of the floral bagging experiments is the importance of pollination by diurnal insect visitors for longan production. The larger proportion numbers of fruits retained on thyrses that received the open and diurnal (i.e., flowers

174

bagged at night) pollination treatments in comparison to other treatments was striking.

Only a few fruits were obtained from flowers on thyrses receiving the wind pollination treatment. Winds were never strong during the period of longan flowering and may have been too weak or at too low a frequency to result in pollination. Similarly, the few fruits obtained in the nocturnal pollination treatment suggest that nocturnal insect visitors were insignificant, either because they were few in number or because they were not efficient pollinators. Moderate numbers of moths and cockroaches visited flowers at night in 2010 but, based on how they fed at flowers without contacting stigmas, they appeared to be nectar robbers rather than pollinators. The bags used to exclude insects for the nocturnal and diurnal pollination treatments did not exclude either wind or self-pollination. The very similar results obtained for nocturnal and wind pollination also suggest that nocturnal visitors did not enhance fruit set.

Self-pollination was of minor importance: fruits were aborted from the self- pollination treatments continuously throughout the entire period of the study, and very few remained at the time of the last check. Longan is reported to be a self-incompatible species (Crane and Walker 1984; Sarfaraz 2011) that requires cross-pollination.

Information on self-incompatibility of longans is scarce in the literature. It is interesting that a few fruits did result from flowers bagged to prevent any type of pollination other than self-pollination, suggesting either in complete self-incompatibility or some pollination of flowers within the paper bags. Unfortunately, the size and quality of these fruits could not be compared with fruits developing from flowers that were effectively pollinated.

175

Longan trees have temporal sexual phases that occur in complex patterns. At any point in time, individual longan trees are usually either in male phase or female phase, with little overlap (Wong 2000). However, in contrast to true duodichogamy (male: female: male), some trees initiated flowering with female flowers and a few trees initiated flowering with flowers of both sexes. Even when male and female flowers occured simultaneously on a tree, usually individual thyrses at a given time were of just one sex.

The complex pattern of floral sex of longans warrants further study. This system of sexual expression requires an external pollen vector, limits self-pollination, and therefore reduces the potential for self-fertilization. However, recent studies suggest that many plant species with patterns of floral sex that favour outcrossing (e.g. dichogamy) are already self-incompatible. It has been suggested that having male flowers and female flowers on different trees or at different times on trees allows plants to export pollen more efficiently (Barrett and Harder 1996; Harder and Barrett 1996).

Several other fruit species in the family Sapindaceae have similar floral traits and pollination requirements (Subhadrabandhu 1990, cited by Davenport and Stern 2005).

Litchi (Litchi chinensis), a species closely related to longan, is reported to be highly self- sterile; wind can contribute to pollination, but insect pollination is stated to be important for crop production (Pandey and Yadava 1970, Davenport and Stern 2005). Honeybee pollination accounts for 39-62% (Anonymous 1981) of litchi fruit yields. Another edible fruit in the Sapindaceae family is rambutan (Nephelium lappacerum). Rambutan has either male or bisexual flowers and pollination is carried out by insects (Lim 1984).

Honeybees augment fruit yield of rambutan by as much as 100% (Manning 2006).

Longan flowers were visited by a high diversity of insects, with 41 families of

176

insects represented in captured specimens. However, relatively few families were common. The most abundant were Apidae (A. cerana and A. mellifera), Syrphidae

(Eristalinus megacephalus and E. lathyrophthalinus), and Calliphoridae (Chrysomya spp.). Numbers of honeybees, by far the most abundant floral visitors, depended on numbers of beehives in the region and their distance to the orchards.

Honeybee visits differed temporally during the day. Both A. cerana and A. mellifera had the same approximate pattern of visitation in 2010: after relatively low numbers of visits in the morning, they increased greatly in the early afternoon, then declined in numbers in the late afternoon (Fig. 5.6a). The temporal pattern for A. cerana in 2011 differed, with numbers of bees visiting flowers peaking in the morning, then gradually declining over the afternoon (Fig. 5.6b). The reasons for this difference are not clear. However, it was much drier during the flowering period in 2011 than in 2010.

Duration of visits by A. mellifera on flowers was not quantified. However, A. mellifera was common in longan orchards in 2010, exhibited behaviour on flowers very similar to A. cerana, and has been shown elsewhere to be an important pollinator of longan (Waite and Hwang 2002, Manning 2006).

The results show that honeybees, A. cerana, were the most frequent floral visitors.

The short duration of each honeybee (A. cerana) visit coupled with the high proportion of honeybees with pollen, the high proportion of that pollen that was longan pollen, and large quantities of pollen on their bodies resulted in A. cerana having a very high relative index of pollinating efficacy. Our results are similar to those reported by Janick and Paull

(2008, cited by Sarfaraz 2011), that bees comprise 98-99% of insects that pollinate

177

longan flowers. The hairy bodies of Apis sp. and their physical contact with anthers and stigmas during foraging also enhance their effectiveness as pollen-transfer vectors.

In our experiments, the other most important pollinators were flies. Calliphorids, predominantly the relatively large and hairy Chrysomya species, visited flowers approximately as quickly as honeybees. Only a moderate proportion of them had longan pollen on their bodies in comparison with flower flies, but during scan sampling they were more frequently observed than flower flies. Flower flies (Syrphidae), particularly those in the genus Eristalinus, also had potential as longan pollinators due to their relatively large hairy bodies. A high proportion (75%) of them carried pollen, all of which was from longan flowers. Syrphids spent about 8 sec on each flower, not significantly different from honeybees. However, they were scarce in the scan samples of floral visitors, only comprising 5% in 2010 and 0.2% of floral visits in 2011, perhaps because the habitat was not appropriate for larval development. House flies (Musca spp.) were slightly more common than flower flies (6% in 2010 and 3% in 2011 of visitors), but very few of them had pollen on their bodies and they remained on flowers for long amounts of time (33 sec), resulting in them having the lowest relative pollination index.

An index of pollinating efficacy for moths was not estimated because data on pollen loads on their bodies and the duration of their visits to flowers were not available.

As discussed previously, while visiting flowers, they rarely contacted stigmas, suggesting they are insignificant as longan pollinators. In addition, the low numbers of fruits obtained from the nocturnal pollination treatment in 2011 also suggest that moths and other night visitors to flowers are not significant pollinators; that conclusion is tentative

178

because the numbers of all floral visitors other than honey bees that year were very low compared to 2010.

The current study suggests that longan has several modes of pollination: diurnal insect pollination is extremely important, with minor contributions from self-pollination and wind pollination. Honeybees were by far the most abundant diurnal insect visitors and contributed greatly to longan yields. The relative proportion of fruit set resulting from diurnal pollination (~84%) was much higher than for wind pollination (8.4%) or self-pollination (7.7%). With their high index of pollinating efficacy, honeybees accounted for ~67% of the total longan yields in 2011. These results agree with studies by Wongsiri (cited by Crane and Walker 1984), Waite and Hwang (2002), and Manning

(2006). Longan growers should recognize the contribution of honeybees and other diurnal insects in their production practices. The use of pesticides to control crop pests during longan flowering has the potential to greatly reduce their crop yields as a result of their potential impact on diurnal insect pollinators, especially honeybees.

179

CHAPTER 6. DISCUSSION AND CONCLUSIONS

This study provided an overview on the importance of honeybees to crop pollination in Vietnam, floral biology of jujube, and pollination biology of jujube and longan. The results indicate that the honeybee, A. cerana, is an important pollinator for jujube and longan production. Future considerations in pollination research in Vietnam are also discussed.

In Chapter 2, based on results of research from other countries, pollination requirements for 39 crops in Vietnam were reviewed, of which 12 crops were discussed in detail. Based on available data, production and economic values of eight highly important crops in Vietnamese agriculture were estimated. The contribution of honeybees as pollinators was almost 50% – ~900 million USD – of the total values of those eight crops. This analysis indicates that the pollination service provided by honeybees and other insects is enormous.

Honeybees are important pollinators for many reasons. They live in large colonies of thousands of worker bees. Bees from a colony can visit many floral species at the same time, yet on individual foraging trips they display floral constancy with pollen and/or nectar usually collected from one single species of flowering plant during a foraging trip.

Hives can be easily moved in and out of crops to meet the specific demands of the flowering crop (Morse and Calderone 2000, Amaya-Márquez 2009). Costs of bee colonies in Vietnam are relatively low in comparison to the value of increased yields of crops that benefit from pollination. Also, in situations in which natural pollinators are present in low numbers, such as in the jujube orchards studied in 2010 and in the longan

180

orchards studied in 2011, the presence of honeybees is critical for high yields and profitability of these crops.

The floral biology and pollination biology of jujube were studied in detail

(Chapters 3 and 4 respectively). Number of flowers per cluster varied between cultivars, from 22 in the F8 cultivar to 27 in the F12 cultivar, and even more for the DM cultivar in two different orchards (i.e., 36-47). These numbers are considerably higher than reported in previous studies: 9.5 flowers per cluster in “Bomple” cultivar in Thailand (Wanichkul and Noppapun 2009), 16-28 flowers (Josan et al. 1980, cited by Pareek et al. 2007) and

10-14 flowers in India (Garhwal 1997, cited by Pareek et al. 2007). In the present study, the continual removal of all open flowers throughout the extended flowering period may have resulted in more accurate counts than were obtained in previous studies.

Alternatively, continual removal of flowers from three clusters per branch may have stimulated those branches to produce more flowers than if flowers had been left intact.

The number of flowers per cluster in jujubes of the “Dao Muon” cultivar differed between the two “Dao Muon” orchards. This difference could be a result of factor(s) that differed between the orchards, including: irrigation regime; moisture retention in the sandy soil; soil nutrients including N, P, K, Ca, Mg, and S; applications of fertilizers; and age of rootstocks. Further studies are needed to better determine the influences of cultural practices on quantity of flowers per cluster and fruit yields.

Ziziphus mauritiana (common names of jujube and ber) had distinct male and female phases of flowering. The development of stamens and dehiscence of anthers before maturation of the stigma indicate that this species is protandrous, as mentioned by

Bertin and Newman (1993). Also, at the beginning of the female phase, when the stigma

181

becomes receptive and flowers secrete nectar, the stamens have drooped onto the petals and anthers are brown and lack pollen. Similar sequential sexual phases have been described for the related species, Z. mucronota (Zietsman and Botha 1992). With this temporal change of floral sex, self-pollination is not possible. Jujube requires vectors to move pollen from male phase day-one flowers to stigmas of day-two flowers in female phase. Most pollination may occur in the late afternoon after anthers have dehisced and pollen is available, although pollination may also occur earlier in the day because dehiscence occurs earlier in the day (~10:00h) in sour jujubes that were planted in the same area. Nectar is a reward for all flower visitors; pollen is an additional reward for bees that require pollen to feed their larvae.

Jujube flowers were visited during the day by insects of many families, particularly honeybees and flies (Chapter 4). By scan sampling and analysis of captured specimens, indices of pollination efficacy of honeybees and three taxa of flies were estimated. These indices took into account abundance, degree of floral constancy, and amount of jujube pollen on the bodies of the most common floral visitors. Apis cerana had a very high index of pollinating efficacy relative to three taxa of flies (flower flies, blow flies, and house flies). In addition, the active movement of honeybees between flowers as they collect pollen further enhances their effectiveness as jujube pollinators.

Floral bagging experiments indicated that only open pollination and diurnal pollination resulted in jujube fruits; nocturnal pollination, self-pollination and wind pollination did not yield fruits. These results indicate that jujube requires diurnal visits from insects for pollination. More specifically, in the year and location of the research, A. cerana was estimated to account for ~97 % of jujube pollination and was therefore a

182

critical pollinator. The bagging studies, coupled with temporal separation of sexual phases in individual jujube flowers, lead to the conclusion that within flower self- pollination is highly unlikely. All results point to A. cerana as the most important pollinator of Z. mauritiana in Vietnam. Unfortunately, because beekeepers always bring large numbers of honeybee colonies to jujube orchards to obtain a honey crop, it was not possible to determine the efficacy of flies as pollinators in the absence of honeybees.

Floral visitation and pollination of longans were studied (Chapter 5). Diversity of flower visiting insects was relatively high in longans. Honeybees were the dominant species on longan flowers, comprising 61% in 2010 and 95% in 2011 of insects recorded during scan sampling. Other diurnal visitors are mostly flies (i.e., flower flies, blow flies, and house flies). Apis cerana had the highest index of pollinating ability (0.86 in 2010;

0.99 in 2011) in comparison with the three other most common taxa of floral visitors. In this thesis, deposition of pollen during single visits by each pollinator species was not quantified. A more complete index of pollination efficacy would include a term describing the efficiency of transfer of pollen from insect bodies to stigmas (Primack and

Silander 1975; Herrera 1987).

Bagging experiments revealed that longan has several modes of pollination: in addition to fruit set resulting from insect visitation to flowers, a few fruits were obtained from wind pollination and self-pollination. Hand crossing between trees did not enhance fruit set compared to open pollination, suggesting that the longan trees studied experienced maximal pollination. Most importantly, the only treatments that had relatively high fruit set were those in which flowers were exposed to pollinators during the day (e.g., open pollination and diurnal pollination treatments). Apis cerana was a

183

critical pollinator. It is estimated that in 2011, their pollination activities resulted in 67% of longan production. The behaviours on flowers and visitation data of the western honeybee, A. mellifera, in 2010 suggest that they are approximately equivalent to A. cerana, but insufficient data were obtained to confirm their relative importance as pollinators. Unfortunately, the number of fruits that were retained on trees up to harvest from the experiments was not sufficient to enable comparisons of the quality of fruits from different pollination treatments.

Jujube and longan have several similarities in pollination biology. They both produce huge numbers of flowers, each of which has a low probability of producing a fruit. They also have similar communities of floral visitors. Apis cerana honeybees make up a large proportion of the total floral visits and they have a very high index of pollination efficacy, contributing > 60% of the fruit yields of jujubes and longans in these studies. Their contribution to the value of the fruit crops is much greater than the value of the honey and other hive products they produce. This result is very important since

Vietnam is an agricultural country, with most people living in rural areas as subsistence farmers. What would happen if their crops failed to set fruits or had low fruit set due to deficiencies of pollinators?

In these studies of jujubes and longans, controlled hand pollination of flowers within trees, between trees, and between cultivars was not conducted. Further research needs to be done to identify the need for inter-tree and inter-varietal pollination for fruit production.

Information compiled in Chapter 2 suggests that many crops in Vietnam require insects for pollination. However, pollination requirements of crops are largely unknown

184

to growers and the values of pollinators for different crops are underestimated or simply not considered. Several reasons may account for this issue. There have been few studies of pollination of crop species in Vietnam. There also is little information on pollination requirements of crops as well as the importance of honeybees and other pollinating insects available through public media. Extension agents are taught only briefly about the role of honeybees and other insects as pollinators of specific crops. Farmers have no way of knowing if they are achieving suboptimal crop yields. On-farm participatory research

(Ahmad and Partap 2008; Veddeler et al. 2008) on pollination may help to overcome some of these deficiencies.

These studies on pollination biology of two fruit crops, jujubes and longans, and for all crops more generally are important for farmers, extension officers, and policy makers, as they highlight the values of honeybees and insects as pollinators. The need to increase food production demands that agriculture become more intensive and more efficient. To control pests, farmers make widespread use of chemical insecticides that can potentially impact insect pollinator populations. There is a need to balance pesticide usage in IPM programs against the need to protect essential insect pollinators. This balance is especially important for honeybees, the hives of which are placed near orchards for honey production. Extension workers can play an important role in enhancing the awareness of farmers about such issues. They act as a bridge between farmers and research institutions to disseminate results of research. Policy makers need to understand the relationship between pollinators and pollination requirements of crops, in order to regulate the use of pesticides and conserve pollinators. In addition, there should be legislation allowing beekeepers to place their beehives near/in orchards for crop

185

pollination as well as for honey production.

This thesis highlights the need for investment in research to identify the diversity of pollinators in rural communities, the pollinators that visit specific crops, the pollination requirements of those crops, and best management practices for pollinators in agroecosystems. Important crops in Vietnamese agriculture should be studied. Particular gaps in knowledge include the pollination requirements of dragon fruits, litchi, and cashew. Such studies will be very important in identifying the role of pollinators in the broad agricultural landscape. They will also be important in raising the awareness of people who work in the field of agriculture, from government personnel to beekeepers and farmers, about the value of honeybees and other insects as pollinators in Vietnamese agriculture.

In conclusion, the contribution of honeybees and other insect pollinators in

Vietnamese agriculture is enormous and significant. Of some major crops reviewed (i.e., avocado, coffee, cucumber, litchi, longan, melon, pumpkin, and sesame), honeybees contribute ~50% of the value of their total yields. The studies reported here demonstrate conclusively that A. cerana is the principal pollinator of jujubes and longans in current

Vietnamese cropping systems. Floral visits of honeybees account for ~97% and ~67% of jujube and longan production, respectively. This research is a call for all Vietnamese who work in the field of agriculture to protect honeybees and other beneficial insects that contribute extensively to agricultural productivity of Vietnam.

186

REFERENCES

Abak, K., Sari, N., Paksoy, M., Kaftanoðlu, O. and Yeninar, H. 1995. Efficiency of bumblebees on the yield and quality of eggplant and tomato grown in unheated glasshouses. Pp. 268-274 In Proceeding of the First International Symposium on

Solanaceae for Fresh Market, Málaga, Spain, 28-31 Mar 1995. Acta Horticulturae Book

No. 412.

Abel, C.A., Wilson, R.L. and Luhman, R.L. 2003. Pollinating efficacy of Osmia cornifrons and Osmia lignaria subsp lignaria (Hymenoptera: Megachilidae) on three Brassicaceae species grown under field cages. Journal of Entomological Sciences 38 (4): 545-552.

Abrol, D.P. 1997. Impact of insect pollination on carrot seed production. Insect

Environment 3 (3): 61.

Abrol, D.P. 2006a. Diversity of pollinating insects visiting litchi flowers (Litchi chinensis

Sonn.) and path analysis of environmental factors influencing foraging behaviour of four honeybee species. Journal of Apicultural Research 45 (4): 180-187.

Abrol, D.P. 2006b. Factors influencing flight activity of Apis florea F., an important pollinator of Daucus carota L. Journal of Apicultural Research 45 (2): 2-6.

Acevedo-Rodriguez, P., Van-Welzen, P.C., Adema F., and Van Der-Ham, R.W.J.M.

2011. Sapindaceae in Kubitzki, K. (Ed) Flowering Plants, Eudicots: Sapindales,

Cucurbitales, Myrtaceae: The Families and Genera of Vascular Plants. Springer-Verlag

Berlin, Germany, pp. 357-407.

18 7

Ackerman, W.L. 1961. Flowering, pollination, self-sterility and seed development of

Chinese jujube. Proceedings of the American Society for Horticultural Science 77: 265-

269. (Only the abstract seen).

Afik, O., Dag, A., Kerem, Z. and Shafir, S. 2006. Analyses of avocado (Persea americana) nectar properties and their perception by honeybees (Apis mellifera). Journal of Chemical Ecology 32 (9): 1949-1963.

Afik, O., Dag, A. and Shafir, S. 2007. Perception of avocado bloom (Lauraceae: Persea americana) by the honeybee (Hymenoptera: Apidae: Apis mellifera). Entomologia

Generalis 30 (2): 135-153.

Ahmad, F., and Partap, U. 2008. Living with honeybees - a decade of enrichment of thought and understanding. Newsletter of the International Centre for Integrated

Mountain Development 2008 (54): 26-30.

Ahmad, R. 1992. Effect of honeybee pollination on fruit yield of cucumber crop. Journal of Zoology 24 (1): 88-90. (Only the abstract seen).

Ahmed, H.M.H., Siddig, M.A. and El-Sarrag, M.S.A. 1989. Honeybee pollination of some cultivated crops in Sudan. Proceedings of the Fourth International Conference on

Apiculture in Tropical Climates, Cairo, Egypt, 6-10 Nov 1988, pp. 100-108.

Alam, M.Z., Quadir, M.A. and Ali, M.S. 1987a. Effect of bee pollination on seed production of carrot, Daucus carota L. Bangladesh Journal of Botany 16 (2): 199-201.

188

Alam, M.Z., Quadir, M.A. and Ali, M. 1987b. Pollinating behaviour of honeybee, Apis cerana F. and its influence on seed production of cauliflower. Bangladesh Horticulture 15

(1): 25-30.

Alam, M.Z., Rahman, M.S. and Ali, M. 1995. Role of pollinating agents on the mode of pollination in mustard. Bangladesh Journal of Botany 24 (1): 13-18.

Allan, P. 1963. Pollination of papaws. Farming in South Africa. 38 (11): 13-15, cited by

McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States

Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Alves, J.E., and Freitas, B.M. 2006. Foraging behavior and pollination efficiency of five bee species on guava (Psidium guajava L.) flowers. Revista Ciência Agronômica 37 (2):

216-220.

Amaral, E. 1972a. Insect pollination of Coffea arabica, and the range of pollen collection by Apis mellifera in flowering coffee plantations. Thesis, University of de Sao Paulo,

Piracicaba, S.P., Brazil. 82 pp.

Amaral, E. 1972b. Production of coffee in the presence and absence of pollinating insects.

2nd Congresso Brazileiro de Apicultura. 1972, pp. 59-67. (Only the abstract seen).

Amaya-Márquez, M. 2009. Floral constancy in bees: A revision of theories and a comparison with other pollinators. Revista Colombiana de Entomología. 35 (2): 206-216.

Anderson, D.L., Sedgley, M., Short, J.R.T. and Allwood, A.J. 1982. Insect pollination of mango in northern Australia. Australian Journal of Agricultural Research. 33 (3): 541-548.

189

Anonymous, n.d. Ziziphus jujuba var. “inermis”. [online]. Available from http://www.learn2grow.com/plants/ziziphus-jujuba-var-inermis/ [accessed 3 Apr 2012].

Anonymous. 1981. Work on bee pollination in India. Indian Bee Journal. 43 (4): 140-144.

Anonymous. 2006. Sap co giong ca chua dat nang suat 100 tan/ha (Tomato variety yielded 100 tons/ha will be available). [online]. Available fromhttp://sokhcn.baria- vungtau.gov.vn/kimlong/news_detail.asp?cat_id=24andnews_id=405[accessed 2 Nov

2008].

Anonymous. 2007a. Hai Duong phat huy uu the rau qua sach (Hai Duong takes an advantages of safety vegetables) [online]. Available fromhttp://www.rauhoaquavietnam.vn/default.aspx?ID=57andLangID=1andtabID=5and

NewsID=3686 [accessed 20 Nov 2008].

Anonymous. 2007b. De thanh long Binh Thuan vuon xa (For dragon fruits in Binh Thuan to be available widely). [online]. Available from http://www.caygiong.com/index.php?Module=ContentandAction=viewandid=342

[accessed 20 Feb 2012].

Anonymous. 2008a. Pollination Canada [online]. Available from http://www.pollinationcanada.ca [accessed 30 Nov 2008].

Anonymous. 2008b. Ha Tinh trong dua hau (Watermelon cultivation in Ha Tinh)

[online]. Available from

190

http://www.rauhoaquavietnam.vn/default.aspx?tabID=3andID=15andLangID=1

[accessed 20 Nov 2008].

Anonymous 2008c. Nong nghiep (Agriculture) [online]. Available from www.baohungyen.vn/content/viewer.asp?a=1729andz=63 - 28k. [accessed 19 Oct 2008].

Araneda, D.X., Breve U.R., Aguilera C.J., Lavin C.J. and Toneatti B.M. 2010. Evaluation of yield component traits of honeybee-pollinated (Apis mellifera L.) rapseed canola

(Brassica napus L.). Chilean Journal of Agricultural Research. 70 (2): 309-314.

Asiko, A.G. 2004. The effect of total visitation time and number of visits by pollinators

(Plebeia sp. and Apis mellifera mellifera) on the strawberry. Thesis, Utrecht University,

The Netherlands, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie 37 (2): 293-315.

Avilán, L., and Rodríguez, M. 1995. Flowering time and avocado yield in the northern region of Venezuela. Agronomía Tropical (Maracay) 45 (1): 35-49.

Bạch T. 2008. Mùa táo ở một vùng quê ven sông Dinh (Jujube production in a village near Dinh river) [online]. Available from http://agriviet.com/?comp=tcnewsandid=3941

[accessed 1 Nov 2008].

Badilla, F., and Ramírez B., W. 1991. Pollination of coffee by Apis mellifera and other insects in Costa Rica. Turrialba 41 (3): 285-288. (Only the abstract seen).

Banda, H.E.J. 1990. Pollination of greenhouse tomatoes by honeybees and bumblebees.

Thesis, University of Wales, College of Cardiff, Cardiff, UK. x + 236 pp.

191

Barrett, S.C.H. and Harder, L.D. 1996. Ecology and evolution of plant mating. Trends in

Ecology and Evolution 11: 73-79.

Baswana, K.S. 1984. Role of insect pollinators on seed production in coriander and fennel. South Indian Horticulture 32 (2): 117-118.

Batra, S.W.T. 1997. Fruit-pollinating bees of the Garhwal Himalaya, U.P., India.

Proceedings of the 7th International Symposium on Pollination. Pollination: From Theory to Practise, Lethbridge, Alberta, Canada, 23-28 Jun 1996. Acta Horticulturae: 437, pp.

325-328.

Bawa, K.S., and Webb, C.J. 1984. Flower, fruit and seed abortion in tropical forest trees - implications for the evolution of paternal and maternal reproductive patterns. American

Journal of Botany 71 (5): 736-751.

Baydar, H., and Gürel, F. 1999. The effects of honeybees on cross-pollination and hybrid seed production in sesame (Sesamum indicum L.). Turkish Journal of Field Crops 4 (1):

21-24.

Berger, L.A., Vaissiere, B.E., Moffett, J.O. and Merritt, S.J. 1988. Bombus spp.

(Hymenoptera, Apidae) as pollinators of male-sterile upland cotton on the Texas high- plains. Environmental Entomology 17 (5): 789-794.

Bertin, R.I., and Newman, C.M. 1993. Dichogamy in angiosperms. Botanical Annual

Review 59 (2): 112-152.

192

Bhatia, R., Gupta, D., Chandel, J.S. and Sharma, N.K. 1995. Relative abundance of insect visitors on flowers of major subtropical fruits in Himachal Pradesh and their effect on fruit set. Indian Journal Agricultural Science 65 (12): 907-912. (Only the abstract seen).

Bhattacharya, A. 2004. Flower visitors and fruit set of Anacardium occidentale. Annales

Botanici Fennici 41 (6): 385-392.

Biao, L., Chang, S., Kun, X., KeXin, Z., XiaoGang, L., DouDou, L., YangPing, Z. and

ChongRen, X. 2009. The oral toxicity of the transgenic Bt + CpTI cotton pollen to honeybees (Apis mellifera). Ecotoxicology and Environmental Safety 72 (4): 1163-1169.

Blair, A.C., and Wolfe, L.M. 2007. The association between floral longevity and pollen removal, pollen receipt, and fruit production in flame azalea (Rhododendron calendulaceum). Canadian Journal of Botany 85 (4): 414-419.

Blanche, K.R., Hughes, M., Ludwig, J.A. and Cunningham, S.A. 2006a. Do flower- tripping bees enhance yields in peanut varieties grown in north Queensland? Australian

Journal of Experimental Agriculture 46 (11): 1529-1534.

Blanche, K.R., Ludwig, J.A. and Cunningham, S.A. 2006b. Proximity to rainforest enhances pollination and fruit set in orchards. Journal of Applied Ecology 43 (6): 1182-

1187.

Bohart, G.E., Nye, W.P., and Hawthorn, L.R. 1970. Onion pollination as affected by different levels of pollinator activity. Utah Agr. Expt. Sta. Bul.. 482, 60pp, cited by

193

McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States

Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Boonithee, A., Juntawong, N., Pechhacker, H. and Hüttinger, E. 1991. Floral visits to select crops by four Apis species and Trigona sp. in Thailand. The Sixth International

Symposium on Pollination, Tilburg, Netherlands, 27-31 Aug 1990, pp. 74-80.

Bouchard, J.N., and Purdie, D.A. 2011. Effect of elevated temperature, darkness, and hydrogen peroxide treatment on oxidative stress and cell death in the bloom-forming toxic cyanobacterium Microcystis aeruginosa. Journal of Phycology 47 (6): 1316-1325.

Bozbek, T., Ozbek, N., Sezener, V., Erdogan, O., Yavas, I. and Unay, A. 2008. Natural crossing and isolation distance between cotton genotypes in Turkey. Scientia Agricola 65

(3): 314-317.

Brewer, J. W. (1974). Pollination requirements for watermelon seed production. Journal of Apicultural Research 13 (3): 207-212.

Buchmann, S.L., and Hurley, J.P. 1978. A biophysical model for buzz pollination in angiosperms. Journal of Theoretical Biology 72 (4): 639-658.

Bumrungsri, S., Sripaoraya, E., Chongsiri, T., Sridith, K. and Racey, P.A. 2009. The pollination ecology of durian (Durio zibethinus, Bombacaceae) in southern Thailand.

Journal of Tropical Ecology 25: 85-92.

Butcher, F.G. 1956. Bees pollinate Litchi blooms. Proceedings of the Florida Litchi

Growers' Association 3: 59-60.

194

Calderone, N.W. 2012. Insect pollinated crops, insect pollinators and US Agriculture:

Trend analysis of aggregate data for the period 1992–2009. PLoS ONE 7 (5): e37235. doi:10.1371/journal.pone.0037235.

Calzoni, G.L. and Speranza, A. 1998. Insect controlled pollination in Japanese plum

(Prunus salicina Lindl.). Scientia Horticulturae 72 (3): 227-237.

Canadian Association of Professional Apiculturists. 1995. A Guide to Managing Bees for

Crop Pollination. Esstee Graphics Inc., Canada.

Cane, J. H. 2005. Squash pollinators of the Americas survey (SPAS) [online]. Available from http://www.ars.usda.gov/Research/docs.htm?docid=12041 [accessed 18 May 2012].

Can-Alonzo, C., Quezada-Euan, J.J.G., Xiu-Ancona, P., Moo-Valle, H., Valdovinos-

Nunez, G.R. and Medina-Peralta, S. 2005. Pollination of “Criollo” avocados (Persea americana) and the behaviour of associated bees in subtropical Mexico. Journal of

Apicultural Research 44 (1): 3-8.

Carman, H. 2012. The estimated impact of bee colony collapse disorder on almond pollination fees [online]. Available from http://giannini.ucop.edu/media/are- update/files/articles/v14n5_4.pdf [accessed 26 Mar 2012].

Carvalho and Krug. 1996. Coffee. Pp. 162–164 In Agriculture Research Service (Ed.)

Agricultural Handbook. United States Department of Agriculture, Washington, DC, cited by Marco Júnior, P.D., and Coelho, F.M. 2004. Services performed by the ecosystem:

195

Forest remnants influence agricultural cultures' pollination and production. Biodiversity

Conservation 13 (7): 1245-1255.

Cauich, O., Quezada-Euán, J.J.G., Macias-Macias, J.O., Reyes-Oregel, V., Medina-

Peralta, S., and Parra-Tabla V. 2004. Behaviour and pollination efficiency of

Nannotrigona perilampoides (Hymenoptera: Meliponini) on greenhouse tomatoes

(Lycopersicum esculentum) in Subtropical México. Horticultural Entomology 97: 475–

481, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie 37 (2): 293-315.

Cervancia, C.R., and Bergonia, E.A. 1991. Insect pollination of cucumber (Cucumis sativus L.) in the Philippines. The Sixth International Symposium on Pollination, Tilburg,

Netherlands, 27-31 Aug 1990,pp. 278-282.

Chacoff, N.P., and Aizen, M.A. 2006. Edge effects on flower-visiting insects in grapefruit plantations bordering premontane subtropical forest. Journal of Applied

Ecology 43 (1): 18-27.

Chacoff, N.P., and Aizen, M.A. 2007. Pollination requirements of pigmented grapefruit

(Citrus paradisi Macf.) from northwestern Argentina. Crop Science 47 (3): 1143-1150.

Chandler, L.D., and Cocke, J. Jr. 1981. Effects of honeybees on cantaloupe yield in the lower Rio Grande Valley of Texas. Southwestern Entomologist 6 (3): 233-236. (Only the abstract seen).

196

Chao, C.C.T. 2005. Pollination study of mandarins and the effect on seediness and fruit size: Implications for seedless mandarin production. HortScience 40 (2): 362-365.

Chen, C.T., and Hsieh, F.K. 2001. Effect of honeybee pollination on the yield and fruit quality of strawberry variety “Toyonoka” (Fragaria xananassa Duch.). Plant Protection

Bulletin (Taipei) 43 (2): 117-127.

Chiari, W.C., de Toledo, V.D.A., Ruvolo-Takasusuki, M.C.C., de Oliveira, A.J.B.,

Sakaguti, E.S., Attencia, V.M., Costa, F.M. and Mitsui, M.H. 2005. Pollination of soybean (Glycine max L. Merril) by honeybees (Apis mellifera L.). Brazilian Archives of

Biology and Technology 48 (1): 31-36.

Chinh, T.X., Sommeijer, M. J., Boot, W. J., and Michener, C. D. 2005. Nest and colony characteristics of three stingless bee species in Vietnam with the first description of the nest of Lisotrigona carpenteri (Hymenoptera: Apidae: Meliponini). Journal of the Kansas

Entomological Society 78 (4): 363-372.

Crane, E., and Walker, P. 1983. The Impact of Pest Management on Bees and

Pollination. Tropical Development and Research Institute, London.

Crane, E., and Walker, P. 1984. Pollination Directory for World Crops. International Bee

Research Association, Gerrards Cross, Bucks, UK.

Cribb, D. 1990. Pollination of tomato crops by honeybees. Bee Craft 72 (8): 228-231.

Cribb, D. 1994. Using the honeybee to pollinate glasshouse tomatoes. Beekeepers'

Quarterly 38: 19-21.

197

Cribb, D.M., Hand, D.W. and Edmondson, R.N. 1993. A comparative study of the effects of using the honeybee as a pollinating agent of glasshouse tomato. Journal of

Horticultural Science 68 (1): 79-88.

Cruz, D. de O., Freitas, B.M., da Silva, L.A., da Silva, S.E.M., Bomfim, I.G.A. 2004. Use of the stingless bee Melipona subnitida to pollinate sweet pepper (Capsicum annuum L.) flowers in greenhouse, Proceeding of the 8th IBRA International Conference. Tropical

Bees andVI Encontro sobre Abelhas, p. 661, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie. 37 (2): 293-315.

Cruz, D.D., Freitas, B.M., da Silva, L.A., da Silva, E.M.S. and Bomfim, I.G.A. 2005.

Pollination efficiency of the stingless bee Melipona subnitida on greenhouse sweet pepper. Pesquisa Agropecuária Brazileira 40 (12): 1197-1201.

Dafni, A., Husband, B.C. and Kevan, P.G. 2005. Practical Pollination Biology.

Enviroquest, Ltd., Cambridge, Ontario, Canada.

Dag, A., and Gazit, S. 2001. Mango pollinators in Israel. Journal of Applied Horticulture

(Lucknow) 2 (1): 39-43. (Only the abstract seen).

Dag, A., and Kammer, Y. 2001. Comparison between the effectiveness of honeybee (Apis mellifera) and bumblebee (Bombus terrestris) as pollinators of greenhouse sweet pepper

(Capsicum annuum). American Bee Journal 141 (6): 447-448.

198

Dag, A., and Mizrahi, Y. 2005. Effect of pollination method on fruit set and fruit characteristics in the vine cactus Selenicereus megalanthus "Yellow pitaya". Journal of

Horticultural Science Biotechnology 80 (5): 618-622.

Dag, A., Zvieli, Y., Afik, O. and Elkind, Y. 2007. Honeybee pollination affects fruit characteristics of sweet pepper grown under net-houses. International Journal of

Vegetable Science 13 (4): 45-59.

D'Albore, G.R. 1984. Observations on the insect pollinators of some Cruciferae (Sinapis alba L., Brassica napus L. subsp. oleifera DC., Brassica oleracea L. v. “Botrytis L.”, and

Brassica rapa L.) cultivated in a specialized area. Apicoltore Moderno 75 (6): 257-264.

Daniel, L.K.P., Loo, Z.X. and Chandran, S. 2010. Obtaining genomic DNA for construction of genomic library. Pp. 549-554 In Southeast Asia Symposium on Quality and Safety of Fresh and Fresh-Cut Produce, Bangkok, Thailand, October 2010. ISHS

Acta Horticulturae Book No. 875.

Das, C.S. and Choudhury, R. 1958. Floral biology of litchi (Litchi chinensis Sonn.).

Indian Horticulture 6 (1): 17-22, cited by McGregor, S.E. 1976. Insect Pollination of

Cultivated Crop Plants. United States Department of Agriculture. Agriculture Handbook.

No. 496. Washington D.C.

Davenport, T.L., and Stern, R.A. 2005. Flowering. In Menzel, C.M. and Waite, G.K.

(Eds.), Litchi and Longan: Botany, Production and Uses. CABI, Wallingford, UK, pp.

87-113.

199

Dawkins, M. S. 2007. Observing Animal Behaviour: Design and Analysis of Quantitative

Data. Oxford University Press Inc., New York, United States.

Da Conceicao, E.S., Delabie, J.H.C. and De Costa N.A. 2004. The entomophily of the coconut tree in question: The evaluation of pollen transportation by ants (Hymenoptera:

Formicidae) and bees (Hymenoptera: Apoidea) in inflorescence. Neotropical Entomology

33 (6): 679-683.

Delaplane, K.S., and Mayer, D.F. 2000a. Avocado. Pp. 156-159 In Delaplane, K.S. and

Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

Delaplane, K.S., and Mayer, D.F. 2000b. Cotton. Pp. 215-218 In Delaplane, K.S. and

Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

Delaplane, K.S., and Mayer, D.F. 2000c. Passion fruits. Pp. 126-128 In Delaplane, K.S. and Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

Delaplane, K.S., and Mayer, D.F. 2000d. Plum and prune. Pp. 247-250 In Delaplane,

K.S. and Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

Delaplane, K.S., and Mayer, D.F. 2000e. Strawberry. Pp. 262-265 In Delaplane, K.S. and

Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

Delaplane, K.S., and Mayer, D.F. 2000f. Tomato. Pp. 270-274 In Delaplane, K.S. and

Mayer, D.F. (Eds.), Crop Pollination by Bees. CABI, Wallingford, UK.

200

Desai, U.T., Randhawa, D.B, Wavhal, K.N. 1986. Floral biology of ber. Journal of

Maharashtra Agricultural University 11:76–78, cited by Pareek, S., Mukherjee, S. and

Paliwal, R. 2007. Floral biology of ber - a review. Agricultural Reviews 28 (4): 277-282.

Devi, K.R., Atluri, J.B. and Reddi, C.S. 1989. Pollination ecology of Ziziphus mauritiana

(Rhamnaceae). Proceedings of the Indian Academy of Sciences (Plant Sciences) 99 (3):

223-240.

Dhaliwal, H.S., Srivastava, S. and Adlakha, R.L. 1977. Insect pollination of litchi, Litchi chinensis Sonn., in the valley areas of the Indian Himalayas. Proceedings of the XXVIth

International Apicultural Congress, Adelaide, 1977: 396.

Dhaliwal, J.S., and Bal, J.S. 1998. Floral and pollen studies in ber (Ziziphus mauritiana

Lamk.). Journal of Research Punjab Agricultural University 35 (1): 36-40.

Dimou, M., Taraza, S., Thrasyvoulou, A. and Vasilakakis, M. 2008. Effect of bumblebee pollination on greenhouse strawberry production. Journal of Apicultural Research 47 (2):

99-101.

Dinh, T.Q., and Pham, D.X. 2001. Publications of Vietnam Bee Research 1990-1999.

Printing Enterprise of Bach Khoa Institute, Hanoi, Vietnam.

Dogterom, M.H., Winston, M.L. and Mukai, A. 2000. Effect of pollen load size and source (self, outcross) on seed and fruit production in highbush blueberry cv. “Bluecrop”

(Vaccinium corymbosum; Ericaceae). American Journal of Botany. 87 (11): 1584-1591.

201

Doutriaux, S., Geisler, C., and Shively, G. 2008. Competing for coffee space: development-induced displacement in the Central Highlands of Vietnam. Rural

Sociology 73 (4): 528–554.

Dung, P. 2009. Ky thuat Trong Tao (Jujube Cultivation Techniques) [online]. Available from http://www.ninhthuan.gov.vn/chinhquyen/sonnnt/Pages/Ky-thuat-trong-tao.aspx

[accessed 16 Jan 2012].

Eardley, C.D., and Mansell, M.W. 1993. Preliminary report on the natural occurrence of insect pollinators in an avocado orchard. Yearbook - South African Avocado Growers'

Association. 16: 127-128.

Eardley, C.D., and Mansell, M.W. 1994. Report on the natural occurrence of insect pollinators in an avocado orchard: Second report. Yearbook - South African Avocado

Growers' Association. 17: 117-118.

El-Sarrag, M., Ahmed, H.M. and Siddig, M.A. 1993. Insect pollinators of certain crops in the Sudan and the effect of pollination on seed yield and quality. Journal of King Saud

University, Agricultural Sciences. 5 (2): 253-262.

Ercan, N., and Onus, A.N. 2003. The effects of bumblebees (Bombus terrestris L.) on fruit quality and yield of pepper (Capsicum annuum L.) grown in an unheated greenhouse. Israel Journal of Plant Sciences 51 (4): 275-283.

Faegri, K., and Pijl, L.V.D. 1979. Principles of Pollination Ecology. Pergamon Press,

Oxford. Toronto.

202

Fajardo, A.C., J., Medina, J.R., Opina, O.S. and Cervancia, C.R. 2008. Insect pollinators and floral visitors of mango (Mangifera indica L. cv. “Carabao”). Philippine Agricultural

Scientist 91 (4): 372-382. (Only the abstract seen).

Faria-Junior, L. R. R., Bendini, J. do N., and Barreto, L. M. R. C. 2008. Pollination efficiency of honeybees and entomophilous pollination in sweet pepper var 'Cascadura ikeda'. Bragantia 67 (2): 261-266. (Only the abstract seen).

Forbes, M.F., and Cervancia, C.R. 1993. Foraging behaviour of Apis cerana F. and Apis mellifera L. (Apidae: Hymenoptera) in Majayjay, Laguna. Pollination in Tropics.

Proceedings of the International Symposium on Pollination in Tropics, 8-13 Aug 1993,

Bangalore, India.

Free, J.B. 1976. Insect Pollination of Tropical Crops. Central Association of Beekeepers,

Ilford, England.

Free, J.B. 1993. Insect Pollination of Crops. Academic Press, Harcourt Brace Jovanovich,

London.

Free, J.B., Raw, A. and Williams, I.H. 1975. Pollination of coconut (Cocos nucifera L.) in Jamaica by honeybees and wasps. Applied Animal Ethology 1: 213-223.

Freitas, B.M. 1995. The pollination efficiency of foraging bees on apple (Malus domestica Borkh) and cashew (Anacardium occidentale L.). Thesis, University of Wales

College of Cardiff, UK. xxiii + 196 pp. (Only the abstract seen).

203

Freitas, B.M. 1996. Foraging behaviour and pollination efficiency of the bee species Apis mellifera and Centris (Hemiseiella) tarsata visiting cashew flowers. Anais da XXXIII

Reunião Anual da Sociedade Brazileira de Zootecnia, Fortaleza, 21 a 26 de Julho, 1996, pp. 666-668.

Freitas, B.M., and Alves, J.E. 2008. Effect of number of floral visits by honeybee (Apis mellifera L.) in the pollination of guava (Psidium guajava L.) cv. “Paluma”. Revista

Ciência Agronômica 39 (1): 148-154.

Freitas, B.M., and Paxton, R.J. 1998. A comparison of two pollinators: the introduced honeybee Apis mellifera and an indigenous bee Centris tarsata on cashew Anacardium occidentale in its native range of NE Brazil. Journal of Applied Ecology 35 (1): 109-121.

Frève, A., Oliveira, D. and Gingras, J. 2001. Insect pollination and production in two old varieties of Damas plums (Prunus domestica L.). Pollination: Integrator of Crops and

Native Plant Systems. Proceedings of the Eight International Pollination Symposium,

Mosonmagyaróvár, Hungary.10-14 Jul 2000, pp. 249-251.

Gallai, N., Salles, J.M., Settele, J., Vaissière, B. E. 2008. Economic valuation of the vulnerability of world agriculture confronted with pollinator decline. Ecological

Economics 68 (3): 810-821 doi:10.1016/j.ecolecon.2008.06.014.

Galuszka, H., and Tegrek, K. 1989. Honeybee (Apis mellifera L.) foraging in seed plantations of hybrid carrot (Daucus carota L.). Proceedings of the XXXIst International

Congress of Apiculture, Warsaw, Poland, 19-25 Aug 1987, pp. 334-336.

204

Gamito, L.M., and Malerbo-Souza, D.T. 2006. Floral visitors and fruit production on sweet orange crop (Citrus sinensis L. Osbeck). Acta Scientiarum Animal Sciences 28 (4):

483-488.

Ganapathi, K., and Viraktamath, S. 2006. Foraging activity of honeybees on Bt cotton hybrids at Dharwad, Karnataka. Indian Bee Journal 68 (1): 74-79.

Garhwal, O.P. 1997. Thesis, Rajasthan Agricultural University. Bikener, India, cited by

Pareek, S., Mukherjee, S. and Paliwal, R. 2007. Floral biology of ber - a review.

Agricultural Reviews 28 (4): 277-282.

Gemmillion-Herren, B., and Ochieng, A.O. 2008. Role of native bees and natural habitats in eggplant (Solanum melongena) pollination in Kenya. Agriculture, Ecosystems and

Environment 127 (1): 31-36.

General Statistics Office of Vietnam (Tong cuc Thong ke Viet Nam). 2008. Statistical

Data. Agriculture, Forestry, and Fishery [online]. Available from http://www.gso.gov.vn/default_en.aspx?tabid=469andidmid=3andItemID=7857

[accessed 28 Dec 2008].

General Statistics Office of Vietnam (Tong cuc Thong ke Viet Nam). 2012. Statistical

Data. Agriculture, Forestry, and Fishery [online]. Available from http://www.gso.gov.vn/default.aspx?tabid=390andidmid=3andItemID=11578 [accessed

2May 2012].

205

Ghini, S., Girotti, S., Calzolari, A., Sabatini, A.G., Alessandrini, A., Zeri, L. and Porrini,

C. 2002. Use of honeybees (Apis mellifera L.) as indicators of the presence of the phytopathogenic bacteria Erwinia amylovora. Insect Social Life (4): 69-77.

Godara, N.R. 1980a. Studies on floral biology and comparative behaviour in ber

(Ziziphus mauritiana Lamk.). Thesis, University of Haryana Agriculture, Hissar, India, cited by Pareek, S., Mukherjee, S. and Paliwal, R. 2007. Floral biology of ber - a review.

Agricultural Reviews 28 (4): 277-282.

Godara, N.R. 1980b. Studies on floral biology and comparative behaviour in ber

(Ziziphus mauritiana Lamk.). Thesis, University of Haryana Agriculture, Hissar, India.

(Only the abstract seen).

Gogoi, B., Rahman, A., Rahman, S. and Deka, M.K. 2007. Foraging behaviour and effect of Apis cerana pollination on fruit set and yield of Assam lemon (Citrus limon). Indian

Journal of Agricultural Science 77 (2): 120-122.

Gosek, J., Ruszkowski, A. and Wojdaszka, J. 1995. Food plants and economic importance of Hylaeus species of the subgenus Dentigera Popov (Hymenoptera,

Colletidae). Pszczelnicze Zeszyty Naukowe 39 (2): 273-279.

Goubara, M., and Takasaki, T. 2003. Flower visitors of lettuce under field and enclosure conditions. Applied Entomology and Zoology 38 (4): 571-581.

206

Goubara, M., and Takasaki, T. 2004. Pollination effects of the sweat bee Lasioglossum villosulum trichopse (Hymenoptera: Halictidae) on genic male-sterile lettuce. Applied

Entomology and Zoology 39 (1): 163-169.

Grover, A., and Sinha, S. K. 1985. Senescence of detached leaves in pigeon pea and chick pea: Regulation by developing pods. Physiologia Plantarum 65 (4): 503-507.

Gu, X.L., Li, Y.P., and Liang, W.H. 2008. The survey of longan industry development in China. Chinese Agricultural Science Bulletin 24: 470-474, cited by Wu,

S.H. 2010. Production status of longan in China. Pp. 37-44 In Proceedings of the III

International Symposium on Longan, Lychee, and other Fruit Trees in Sapindaceae

Family, Fuzhou, China, May 2010. ISHS Acta Horticulturae Book No. 863.

Gulati, R., Sharma, S.K. and Chhillar, B.S. 2005. Foraging behaviour and efficiency of

Asian bees in cotton (Gossypium hirsutum L.) in India. Honeybee Science 26 (1): 25-28.

Guseinov, S.I., and Ragim-Zade, M.S. 1985. Pollination of the cotton plant by Caucasian bees. Izvestiya Akademii Nauk Azerbaĭdzhanskoĭ SSR, Biologicheskikh Nauk: 40-46.

(Only the abstract seen).

Guzman-Novoa, E., Sánchez, A., Page, R.E., J. and García, T. 1996. Susceptibility of

European and Africanized honeybees (Apis mellifera L) and their hybrids to Varroa jacobsoni Oud. Apidologie 27 (2): 93-103.

207

Hache, V. and Xu, L.P. 1995. The jujube, a fruit of high potential not only in China [in

German]. Flüssiges Obst 62: 35-36, cited by William, H. 2002. Notes on economic plants. Economic Botany 56 (2): 198-206.

Hammons, R.O. 1963. Artificial cross-pollination of the peanut with bee-collected pollen.

Crop Science 3: 562-563, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated

Crop Plants. United States Department of Agriculture. Agriculture Handbook. No. 496.

Washington D.C.

Hammons, R.O., and Leuck, D.B. 1966. Natural cross-pollination of the peanut, Arachis hypogaea L., in the presence of bees and thrips. Agronomy Journal 58: 396.

Harder, L.D. and Barrett, S.C.H. 1996. Pollen dispersal and mating patterns in animal pollinated plants. Pp. 140-190 In Lloyd. D.G. and Barrett, S.C. H. (Eds). Floral Biology:

Studies on Floral Evolution in Animal-Pollinated Plants. Chapman and Hall: New York.

Haslbachova, H., and Kubisova, S. 1982. The pollinating activity of honeybees on onion

(Allium cepa). Number of visits by bees and its effect on seed yield. Sbornik UVTIZ,

Zahradnictvi 9 (2): 119-126. (Only the abstract seen).

Hawthorn, L.R., Bohart, G.E. and Toole, E.H. 1960. Carrot seed production as affected by insect pollination. Utah Agricultural Experiment Station, p. 18. (Only the abstract seen).

Heard, T.A. 1999. The role of stingless bees in crop pollination. Annual Review of

Entomology 44: 183-206.

208

Heard, T.A., Vithanage, V. and Chacko, E.K. 1990. Pollination biology of cashew in the northern territory of Australia. Australian Journal of Agricultural Research 41 (6): 1101-

1114.

Heide, F.F. R. 1923. Biological observations of Arachis hypogaea L. [Duch East Indies]

Alg. Proefsta. V. Landb. Meded. 14: 1-19, cited by McGregor, S.E. 1976. Insect

Pollination of Cultivated Crop Plants. United States Department of Agriculture.

Agriculture Handbook. No. 496. Washington D.C.

Herrera. C.M. 1987. Components of pollinator "quality": Comparative analysis of a diverse insect assemblage 50 (1): 79-90.

Higo, H.A., Rice, N.D., Winston, M.L. and Lewis, B. 2004. Honeybee (Hymenoptera:

Apidae) distribution and potential for supplementary pollination in commercial tomato greenhouses during winter. Journal of Economic Entomology 97 (2): 163-170.

Hikawa, M., and Miyanaga, R. 2006. Effects of pollination by Melipona quadrifasciata anthidioides on forcing eggplant cultures. Horticultural Research (Japan) 5 (2): 149-152.

Ho, Y.F. 1993. Xylocopa latipes, the Carpenter Bee. Forest Research Institute, Kuala

Lumpur, Malaysia.

Hoang, N. 2006. “Ba cai tao” va “Hai ho tro” cho nong san Lao Cai (“Three improvements” and “Two assistances” for Agricultural products in Lao Cai). Available fromhttp://ipsard.gov.vn/news/newsdetail.asp?targetID=623 [accessed 20 Nov 2008].

209

Hofs, J., Schoeman, A.S. and Pierre, J. 2008. Diversity and abundance of flower-visiting insects in Bt and non-Bt cotton fields of Maputaland (KwaZulu Natal Province, South

Africa). International Journal of Tropical Insect Science 28 (4): 211-219.

Hogendoorn, K., Steen, Z. and Schwarz, M.P. 2000. Native Australian carpenter bees as a potential alternative to introducing bumblebees for tomato pollination in greenhouses.

Journal of Apicultural Research 39 (1-2): 67-74.

Holanda Neto, J. P. D, Freitas, B.M., Bueno, D.M. and Araújo, Z.B.D. 2002. Low seed/nut productivity in cashew (Anacardium occidentale): Effects of self-incompatibility and honeybee (Apis mellifera) foraging behaviour. Journal of Horticultural Science and

Biotechnology 77 (2): 226-231.

Hooper, C.H. 1936. Plums: Notes on their pollination, order of flowering of varieties and insect visitors to the blossoms. Journal South Eastern Agr. Col. 38: 131-140, cited by

McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States

Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Howpage, D., Spooner-Hart, R. and Vithanage, V. 2001. Influence of honeybee (Apis mellifera) on kiwifruit pollination and fruit quality under Australian conditions. New

Zealand Journal of Crop and Horticultural Science 29 (1): 51-59.

Hyung Rae, L., Jeong Wha, K., Eun Jeoung, L. and Seong Ok, L. 1995. Foraging activities and pollination efficacies of pollinators on pepper (Capsicum annuum), perilla

(Perilla frutescens var. “Japonica”) and sesame (Sesame orientale). Korean Journal of

Apiculture 10 (2): 117-122. (Only the abstract seen).

210

Ikeda, F., and Tadauchi, Y. 1995. Use of bumblebees as pollinators on fruit and vegetables. Honeybee Science 16 (2): 49-56.

Ingram, M., Nabhan, G. and Buchmann, S. 1996. Our forgotten pollinators: Protecting the birds and bees[online]. Available from http://www.pmac.net/birdbee.htm [accessed

30 Mar 2012].

Ish-Am, G., and Eisikowitch, D. 1990. The behaviour of honeybees during visits to the avocado flower and their contribution to its pollination. Alon Hanotea 40 (7): 591-606.

Janick, J. and Paull, R.E. 2008. The Encyclopedia of Fruit & Nuts, cited by Sarfaraz, I.

Do Longan Trees Need Cross-Pollination From Another Longan Tree? [online].

Available from http://www.ehow.com/info_10036528_longan-trees-need- crosspollination-another-longan-tree.html. [accessed 30 Apr 2012].

Jay, S.C. 1990. Pollination of crops in Manitoba by bees. Canadian Beekeeping 16 (1):

16-17.

JianDong, A., and WenFeng, C. 2011. Economic value of insect pollination for fruits and vegetables in China. Scientia Agricultura Sinica 54 (4): 443-450.

JinHua, H. 2007. Techniques for increasing fruit set rate in litchi cultivar Chenzi. China

Fruits 5: 66-67. (Only the abstract seen).

Josan, J.S., Jawanda, J.S., Bal, J.S., and Singh, R. 1980. Studies on the floral biology of ber (Ziziphus mauritiana Lamk.). 1. Flowering habit, flower bud development, time and duration of flowering, flowering morphology, time of anthesis and dehiscence. Punjab

211

Horticultural Journal 20 (1 & 2): 156-161, cited by Pareek, S., Mukherjee, S. and Paliwal,

R. 2007. Floral biology of ber - a review. Agricultural Reviews 28 (4): 277-282.

Josan, J.S., Jawanda, J.S., Bal, J.S., and Singh. 1981. Studies on the floral biology of ber.

Pollen study, receptivity of stigma and mode of pollination. Punjab Horticultural Journal

21 (1 & 2): 61-63, cited by Dhaliwal, J.S., and Bal, J.S. 1998. Floral and pollen studies in ber (Ziziphus mauritiana Lamk.). Journal of Research Punjab Agricultural University 35

(1): 36-40.

Jothi, B.D., and Tandon, P.L. 1993. Insect pollinator activity on ber (Zizyphus mauritiana

Lamk.). Pollination in Tropics. Proceedings of the International Symposium on

Pollination in Tropics, 8-13 Aug 1993, Bangalore, India, pp. 210-211.

Kakutani, T., Inoue, T., Tezuka, T. and Maeta, Y. 1993. Pollination of strawberry by the stingless bee, Trigona minangkabau, and the honeybee, Apis mellifera - an experimental study of fertilization efficiency. Researches on Population Ecology 35 (1): 95-111.

Kalman, C. 1979. Avocado pollination. Proceeding IV International Symposium on

Pollination, 1978: 207-209, cited by Crane, E., and Walker, P. 1984. Pollination

Directory for World Crops. International Bee Research Association, Gerrards Cross,

Bucks, UK.

Kato, E.C., and Nogueira-Couto, R.H. 2002. Pollination of winter and net-melon

(Cucumis melo L.). Naturalia (São Paulo) 27: 201-210. (Only the abstract seen).

212

Kearns, C.A., and Inouye, D.W. 1993. Techniques for Pollination Biologists. University

Press of Colorado, Boulder, CO.

KienHock, L., IouZen, C. and TsuLiang, C. 2002. Factors affecting fruitfulness in durian

(Durio zibethinus Murr.). Flowering and pollination. Journal of the Chinese Society for

Horticultural Science 48 (4): 287-297. (Only the abstract seen).

King, J., Exley, E.M. and Vithanage, V. 1989. Insect pollination for yield increases in litchi. Pp. 142-145 In Batten, D.J. (Ed.), Proceedings of the Fourth Australian Conference on Tree and Nut Crops. Exotic Fruit Growers’ Association, Lismore, Australia, cited by

Davenport, T.L., and Stern, R.A. 2005. Flowering. Pp. 87-113 In Menzel, C.M. and

Waite, G.K. (Eds.), Litchi and Longan: Botany, Production and Uses. CABI,

Wallingford, UK, pp. 87-113.

Klein, A.M., Steffan-Dewenter, I. and Tscharntke, T. 2003a. Bee pollination and fruit set of Coffea arabica and C. canephora (Rubiaceae). Americal Journal of Botany 90 (1):

153-157.

Klein, A.M., Steffan-Dewenter, I. and Tscharntke, T. 2003b. Pollination of Coffea canephora in relation to local and regional agroforestry management. Journal of Applied

Ecology 40 (5): 837-845.

Klein, A.M., Vaissière, B.E., Cane, J.H., Steffan-Dewenter, I., Cunningham, S.A.,

Kremen, C. and Tscharntke, T. 2007. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society of London. Series B, Biological

Sciences, 2007. 274 (1608): 303-313.

213

Kolesnik, M.G. 1987. The role of useful insect fauna in producing high carrot seed yields. Journal Nauchno-Tekhnicheskiĭ Byulleten' Vsesoyuznogo Ordena Lenina i

Ordena Druzhby Narodov Nauchno-Issledovatel'. Skogo Instituta Rastenievodstva Imeni

N. I. Vavilova: 62-63. (Only the abstract seen).

Kordakova, Z.M. 1956. Honeybees and pollination of seed plants of the common onion.

Pp. 163-171 In Krishchunas, I.V., and Gubin, A.F. (Eds.), Pollination of agricultural plants, Moskva, Gos. Russia, cited by McGregor, S.E. 1976. Insect Pollination of

Cultivated Crop Plants. United States Department of Agriculture. Agriculture Handbook.

No. 496. Washington D.C.

Kumar, J., and Rao, K.V.K. 1991. Pollinating efficiency of some bee visitors to the carrot

(Daucus carota L.) crop in mid hills of Himachal Pradesh (India). Indian Bee Journal 53

(1-4): 34-38.

Kumar, J., Mishra, R.C. and Gupta, J.K. 1989. Effect of honeybee pollination on onion

(Allium cepa L.) seed production. Indian Bee Journal 51 (1): 3-5.

Kuti, J.O., and Rovelo, C. 1992. Effect of pollination techniques on fruit set and seed yield of egusi melon (Citrullus colocynthis). Report - Cucurbit Genetics Cooperative 15:

80-81. (Only the abstract seen).

Kuvanci, A., Gunbey, B., Konak, F. and Karaoglan, Y. 2010. The effects of honeybees

(Apis mellifera L.) and other insects on pollination of strawberry (Fragaria sp.) plants.

Uludag Aricilik Dergisi 10 (1): 28-34.

214

Langridge, D.F., and Goodman, R.D. 1985. Honeybee pollination of Japanese plums

(Prunus salicina Lindl. cv. “Satsuma”) in the Goulburn Valley, Victoria, Australia.

Australian Journal of Experimental Agriculture 25: 227-230.

Le Thi, T.H., Dinh, T. Q., Hoang, A.T., Trach, L.K., Lövei, G.L., Pires, C.S.S., FangHao,

W., JianYing, G. and Andow, D.A. 2008. Potential effects of transgenic cotton on flower visitors in Vietnam. pp. 236-257 In D.A. Andow, A. Hillbeck and N.V. Tuat (Eds.),

Environmental Risk Assessment of Genetically Modified Organisms: Challenges and

Opportunities with Bt Cotton in Vietnam, CABI, Wallingford, UK.

Lee, H.R., Kim, J.W. and Choi, S.Y. 1988. Foraging activity of honeybees (Apis mellifera) and pollination effects on several crops. Korean Journal of Apiculture 3 (1):

68-80. (Only the abstract seen).

Lemasson, M. 1987. Usefulness of the honeybee (Apis mellifera) in the pollination of glasshouse-grown Gherkin (Cucumis sativus), melon (Cucumis melo), and tomato

(Lycopersicum esculentum). Revue de l'Agriculture 40 (4): 915-924. (Only the abstract seen).

Levin, M.D. 1984. Value of bee pollination to United States agriculture. American Bee

Journal 124 (3): 184-186.

Lim, A.L. 1984. The reproductive biology of rambutan Nephelium lappaceum

(Sapindaceae). Gardens' Bulletin (Singapore) 37 (2): 181-192.

215

Lim, T.K., and Luders, L. 1998. Durian flowering, pollination and incompatibility studies. Annual Applied Biology 132: 151-165, cited by cited by Klein et al. 2007.

Importance of pollinators in changing landscapes for world crops. Proceedings of the

Royal Society of London. Series B, Biological Sciences 274 (1608): 303-313.

Liu, M. J., and Zhao, Z. H. 2008. Germplasm resources and production of jujube in

China. Proceedings of the International Jujube Symposium, Baoding, China, 21-25 Sep

2008. (840). 25-32.

Liu, M.J., Zhao, Z.H. 2009. International survey and discussion on the names of genus jujube and its main species. Proceedings of the International Jujube Symposium,

Baoding, China, 21-25 Sep 2008. (840). 81-88.

Liu, P., Liu, M.J., Zhao, Z.H., Liu, X.Y., Wang, J.R. and Yan, C. 2009. Investigation on the characteristics of fruiting and seed development in Chinese jujube (Ziziphus jujuba

Million.). Proceedings of the International Jujube Symposium, Baoding, China, 21-25

Sept 2008. (840): 209-214.

Lorenzon, M.C.A., Martinho, M.R., Freire, J.A.H. and Reis, F.P. 1993. Pollination effect by honeybees to onion hybrid seed production. Pesquisa Agropecuária Brazileira 28 (2):

229-235. (Only the abstract seen).

Lu, K., Zhang, H.J. and Huang, Y. 2007. The analysis of present production situation on the primary tropical fruits in 2005. World Tropical Agriculture Information 45: 4-10, cited by Wu, S.H. 2010. Production status of longan in China. Pp. 37-44 In Proceedings

216

of the III International Symposium on Longan, Lychee, and other Fruit Trees in

Sapindaceae Family, Fuzhou, China, May 2010. ISHS Acta Horticulturae Book No. 863.

Lyrene, P.M. 1983. Flowering and fruiting of Chinese jujubes in Florida. Horticultural

Science 18 (2): 208-209.

MacDaniels, L.H. 1942. Notes on the pollination of the Italian prune. Amer. Soc. Hort.

Sci. Proc. 40: 84-86, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop

Plants. United States Department of Agriculture. Agriculture Handbook. No. 496.

Washington D.C.

Malagodi-Braga, K.S. and Kleinert, A.M.P. 2004. Could Tetragonisca angustula Latreille

(Apinae, Meliponini) be used as strawberry pollinator in greenhouses? Australian Journal of Agricultural Research 55: 771–773, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie 37 (2): 293-315.

Malaipan, S., Kongpitak, P. and Sujitratanan, S. 1992. Comparative study on the effect of insecticides and neem extracts to the thrips and honeybee pollinators of pomelo cultivar

Tongdee. Kasetsart Journal, Natural Sciences 26 (2): 146-150. (Only the abstract seen).

Malerbo-Souza, D., Toledo, V., Silva, S.R. and Sousa, F.F. 2000. Pollination in avocado flowers (Persea americana Million.). Acta Scientiarum 22 (4): 937-941. (Only the abstract seen).

Malerbo-Souza, D.T. and Nogueira-Couto, R.H. 1997. Atrativo para as abelhas Apis mellifera e polini-zac a o em cafe (Coffea arabica L.). Mensagem Doce 44: 6–11, cited

217

by Marco Júnior, P.D. and Coelho, F.M. 2004. Services performed by the ecosystem:

Forest remnants influence agricultural cultures' pollination and production. Biodiversity

Conservation 13 (7): 1245-1255.

Malerbo-Souza, D.T., and Nogueira-Couto, R.H. 2002. Pollination by insects in three orange (Citrus sinensis L. “Osbeck” cultivars). Científica (Jaboticabal). 30 (1): 79-87.

(Only the abstract seen).

Malerbo-Souza, D.T., Nogueira-Couto, R.H., and Couto, L.A. 2004. Honeybee attractants and pollination in sweet orange, (Citrus sinensis L.“Osbeck” var). Journal of

Venomous Animals and Toxins Including Tropical Diseases. 10 (2): 144-153.

Malerbo-Souza, D.T., Nogueira-Couto, R.H., Couto, L.A. and da Souza, J.C.. 2003.

Attractives to Africanized honeybees and pollination in coffee flowers (Coffea arabica

L.). Brazilian Journal of Veterinary Research and Animal Science 40 (1): 272-278. (Only the abstract seen).

Mann, G.S., and Singh, G. 1983. Activity and abundance of pollinators of plum at

Ludhiana (Punjab). American Bee Journal 123 (8): 595.

Manning, R. 2006. Honeybee pollination: Technical data for potential honeybee- pollinated crops and orchards in Western Australia [online]. Available from http://www.agric.wa.gov.au/PC_91801.html?s=1971607857 [assessed 30 Apr 2012].

218

Marco Júnior, P.D., and Coelho, F.M. 2004. Services performed by the ecosystem: Forest remnants influence agricultural cultures' pollination and production. Biodiversity

Conservation 13 (7): 1245-1255.

Martin, E.C., and McGregor, S.E. 1973. Changing trends in insect pollination of commercial crops. Annual Review of Entomology 18: 207-226.

Matuszak, W. 1995. A solitary bee (Osmia rufa L.) in horticultural seed production.

Materiały ogólnopolskiej konferencji naukowe Nauka Praktyce Ogrodniczej z okazji

XXV-lecia Wydziału Ogrodniczego Akademii Rolniczej w Lublinie. 1995, pp. 911-913.

(Only the abstract seen).

McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States

Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

McGregor, S.E. and Todd, F.E. 1955. Cotton is benefited by bees. Prog. Agr. In Ariz. 7 (1):

12, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United

States Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

McMullen, C.K. 2012. Pollination of the heterostylous Galápagos native, Cordia lutea

(Boraginaceae). Plant Systematics and Evolution 298 (3): 569-579.

Meeuwsen, F.J.A.J. 2000. Stingless bees for pollination purposes in greenhouses. Pp.

143-147 In: Sommeijer, M.J., Ruijter, A. de (Eds.), Insect Pollination in Greenhouses:

Proceeding Specialists’ Meeting Held in Soesterberg, The Netherlands, cited by Slaa et

219

al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie 37

(2): 293-315.

MeiYing, L., ShiJin, O., JunZhi, X., and GuiXiang, H. 2004. Observation on the fruit growth and physiological fruit drop regulation of Dawuyuan longan cultivar. South China

Fruits 33 (4): 28-29.

Meléndez-Ramírez, V., Parra-Tabla, V., Kevan, P.G., Ramírez-Morillo, I., Harries, H.,

Fernández-Barrera, M. and Zizumbo-Villareal, D. 2004. Mixed mating strategies and pollination by insects and wind in coconut palm (Cocos nucifera L. (Arecaceae)):

Importance in production and selection. Agricultural and Forest Entomology 6 (2): 155-163.

Menzel, R. and Shmida, A. 1993. The ecology of flower colours and the natural colour vision of insect pollinators: The Israeli flora as a study case. Biological Review. 68: 81–120.

Meyer, R. 1995. Three new crops with proven market potential [online]. Available from http://www.newcrops.uq.edu.au/acotanc/papers/meyer.htm [accessed 23 May 2012].

Michelbacher, A.E., Smith, R.F., and Hurd, R.D. 1964. Bees are essential pollination of squashes, gourds and pumpkins. Calif. Agr. 18 (5): 2-4, cited by McGregor, S.E. 1976.

Insect Pollination of Cultivated Crop Plants. United States Department of Agriculture.

Agriculture Handbook. No. 496. Washington D.C.

Michener, C.D. 2000. The Bees of the World. Johns Hopkins University Press,

Baltimore.

220

Michener, C.D., McGinley, R.J. and Danforth, B.N. 1994. The Bee Genera of North and

Central America (Hymenoptera: Apoidea). Smithsonian Institution Press, Washington.

Minh, S. 2004. Ong rung giup tang san luong cafe (Honeybee increases coffee production) [online]. Available from http://vietbao.vn/Khoa-hoc/Ong-rung-giup-tang- chat-luong-san-luong-ca-phe/20222900/197/http://giaoduc.edu.vn/news/chuyen-doanh- nghiep-734/mat-ong-viet-nam-bi-tac-o-my-180388.aspx [accessed 15 Mar 2012].

Ministry of Agriculture and Rural Development of Vietnam. 2009. Danh muc bo sung phan bon duoc phep san xuat, kinh doanh va su dung tai vietnam (List of additional fertilizers that are allowed to produce, trade, and use in Vietnam). Available from http://www.chinhphu.vn/portal/page/portal/chinhphu/hethongvanban?class_id=1&_page

=6&mode=detail&document_id=154209 [accessed 20 Feb 2012].

Ministry of Science and Technology. 2009. Gioi thieu khai quat ve tinh Ha Tay (General introduction about Ha Tay) [online]. Available from http://www.most.gov.vn/c_so_khcn/mlfolder.2006-07-05.8541258373/mldocument.

2006-07-06.7368595747/mldocument_view [accessed 13 Mar 2009].

Missouri Botanical Garden. 2009. Biology of Plants: Pollination [online]. Available from http://www.mbgnet.net/bioplants/pollination.html [accessed 3 Oct 2009].

Mitra, S.K., Pereira, L.S., Pathak, P.K. and Majumdar, D. 2005. Fruit abscission pattern of lychee cultivars. Proceedings of the Second International Symposium on Lychee,

Longan, Rambutan and other Sapindaceae Plants, Chiang Mai, Thailand, 25-28 August,

2003: 215-218.

221

Miyamoto, M., Ono, M., Sasaki, M., Abe, H. and Kenmochi, I. 2006. Development of fruiting promotion system using honeybees in semi-forced eggplant culture. Effectiveness of pollination by the European honeybee. Japanese Journal of Applied Entomology and

Zoology 50 (4): 297-304. (Only the abstract seen).

Mizuno, S., Miyazato, K., Yoshida, T., Oku, H. and Kiyokawa, K. 2004. Effect of

Japanese honeybee and blow fly on the pollination and resultant fruit quality of mango

"Irwin". Bulletin of the Faculty of Agriculture Tamagawa University 43-44: 29-38. (Only the abstract seen).

Mizuno, S., Yoshida, T., Kiyokawa, K. and Sasaki, M. 2007. Pattern of visiting flowers and pollination efficiency of three kinds of bees on “Irwin” mango fruits grown in plastic greenhouses. Japanese Journal of Tropical Agriculture 51 (3): 116-122. (Only the abstract seen).

Moffett, J.O., and Rodney, D.R. 1973. Honeybee visits increase yields of “Orlando” tangelo. HortScience 8 (2): 100.

Moffett, J.O., and Rodney, D.R. 1975. Effect of honeybee visits on the yields of caged lemon trees. American Bee Journal 115 (8): 310-311, 313.

Moffett, J.O., Banks, D.J. and Pittman, R.M. 1986. Floral visits by honeybees to three caged peanut genotypes and the resulting increase in hybrid seed. American Bee Journal

126 (12): 833.

222

Morandin, L.A., and Winston, M.L. 2005. Wild bee abundance and seed production in conventional, organic, and genetically modified canola. Ecological Applications 15 (3):

871-881.

Morse R., and Calderone N. 2000. The value of honeybees as pollinators of U.S. crops in

2000 [online]. Available from http://www.beeculture.com/beeculture/pollination2000/pg1.html. [accessed 1 Jan 2009].

Morton, J.M., and Miami, F. 1987. Fruits of warm climates [online]. Available from http://www.hort.purdue.edu/newcrop/morton/longan.html#Varieties [accessed 14 Jan

2009].

Murao R. 2011. Taxonomic notes on Lasioglossum (Lasioglossum) subopacum (Smith) and L. (L.) okinawa Ebmer et Maeta (Hymenoptera, Halictidae) from Asia. ZooKeys 143:

83-92.

Nachappa, M.S., and Viraktamath, S. 2005a. Bee visitation and foraging activity of

Indian honeybee, Apis cerana F., in Bt and non-Bt cotton hybrids at Dharwad. Journal of

Palynology 40: 173-179.

Nachappa, M.S., and Viraktamath, S. 2005b. Rock bee, Apis dorsata F., visitation on Bt and non-Bt cotton hybrids at Dharwad, Karnataka. Journal of Palynology 40: 181-187.

Nadia, T., Machado, I.C. and Lopes, A.V. 2007. Reproductive phenology and pollination system of Ziziphus joazeiro Mart. (Rhamnaceae): The role of Apis mellifera and

223

autochthonous floral visitors as pollinators. Acta Botanica Brazilica 21 (4): 835-845.

(Only the abstract seen).

Nakata, M., Sugiyama, N., and Pankasemsuk, T. 2005. Morphological study of the structure and development of longan inflorescence. Journal of the American Society for

Horticultural Sciences 130 (6): 793-798.

Neeraja, G., Reddy, S.A. and Babu, R.S. 1996. Pollen studies and pollination in ber

(Zizyphus mauritiana Lamk.). Journal of Research, Acharya. N. G. Ranga Agricultural

University 24 (1): 100-105.

Nguyen, C.M. 2008. Trong tao (Cultivation of jujube). Agricultural Publishing House,

Hanoi, Vietnam.

Nguyen, D.V. 2012. Nhan muon Khoai Chau - Dac san nhan long Hung Yen (Khoai

Chau late harvest longan - Special product of "Long" longan in Hung Yen) [online].

Available from http://hungyen.gov.vn/vi- vn/khoaichau/Pages/Article.aspx?ChannelId=13&articleID=147 [accessed 14 April

2012].

Nguyen, L.V. 2008. Thong ke Nong Lam Thuy san 2007 (Statistics of Agriculture,

Forestry, and Fishery of Vietnam; unpublished book).

Nguyen, T. 2009. Lam giau tu carot (Being rich from carrot). [online]. Available from http: //www.baomoi.com/Home/AmThuc/baodientu.chinhphu.vn/Lam-giau-tu-carot/3538016.epi. [accessed 29 Mar 2012].

224

Nguyen, T.Q., and Pham, B.T. 1994. Harmony or conflict: Apis mellifera and Apis cerana in Southern Vietnam. Beekeeping & Development. 32: 5-8.

Nianhe and Gadex. 2012. Sapindaceae. [online]. Available from http: http://flora.huh.harvard.edu/china/mss/volume12/Sapindaceae.pdf. [accessed 29 Apr

2012].

Nicholson, S.W. and Thornburg, R. 2007. Nectar chemistry. Pp. 215-263 In Nicholson,

S.W., M. Nepi, and E. Pacini (Eds.), Nectaries and Nectar. Springer: Dordrecht.

Nikitina, A.I. 1950. Honeybees raise seed yields of turnip and rutabaga. Pchelovodstvo. (5):

271-274, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United

States Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Oanh L., 2009. Tiem nang phat trien cay me tren dat Binh Tan (Potential on sesame development in Binh Tan, Vietnam) [online]. Available from http://www.nongnghiep.vinhlong.gov.vn/Default.aspx?tabid=36andcategoryid=11andite mid=1058 [accessed10 May 2009].

Oldroyd, B.P., and Wongsiri, S. 2006. Asian Honeybees: Biology, Conservation, and

Human Interactions. Harvard University Press, Cambridge, MA, USA.

Olschewski, R., Tscharntke, T., Benítez, P.C., Schwarze, S. and Klein, A.M. 2006.

Economic evaluation of pollination services comparing coffee landscapes in Ecuador and

Indonesia. Ecology and Society.11 (1): article 7.

225

Orlova, K.B., Skrebtsov, M.F. and Skrebtsova, N.D. 1981. Effect of bee pollination on seed production by onions (Allium cepa). Trudy Po Prikladnoĭ Botanike, Genetike i

Selektsii 69 (2): 66-69. (Only the abstract seen).

Ortiz-Sánchez, F.J., and Cabello-García, T. 1991. Use of insects in the pollination of subtropical crops. Agricola Vergel. 10 (119): 692-694. (Only the abstract seen).

O’Toole, C., and Raw, A. 1991. Bees of the World. Blandford Publishing, London, UK.

Pagano, M., and Gauvreau, K. 1993. Principles of Biostatistics. Belmort, CA. Duxbury

Press.

Panda, P., Nanda, U.K., Mohapatra, H. and Padhi, J. 1989. Insect pollination in some oilseed crops in Orissa, India. Indian Bee Journal 51 (3): 97-98.

Panda, P., Sontakke, B.K. and Sarangi, P.K. 1988. Preliminary studies on the effect of bee (Apis cerana Fabr.) pollination on yield of sesamum and niger. Indian Bee Journal 50

(3): 63-64.

Pandey, R.S., and Yadava, R.P.S. 1970. Pollination of litchi (Litchi chinensis) by insects with special reference to honeybees. Journal of Apicultural Research 9: 103-105.

Papademetriou, M.K., and Dent, F.J. 2002. Litchi production in the Asia-Pacific region

[online]. Available from http://www.fao.org/docrep/005/ac684e/ac684e0d.htm#TopOfPage [accessed 21 Nov

2008].

226

Pardede, D., Randriani, E. and Marfuati, A. 1986. The pollination of coconut at the

Pakuwon Sub-Research Institute for coconut. Pemberitaan, Penelitian Tanaman Industri,

Indonesia 12 (1): 26-35. (Only the abstract seen).

Pareek, S., Mukherjee, S. and Paliwal, R. 2007. Floral biology of ber - a review.

Agricultural Reviews 28 (4): 277-282.

Parsons, C. 2012. Bee supply better for 2012 almond pollination [online]. Available fromhttp://westernfarmpress.com/markets/bee-supply-better-2012-almond-pollination

[accessed 26 Mar 2012].

Partap, U. 2000. Pollination of strawberry by the Asian hive bee, Apis cerana. Asian

Bees and Beekeeping: Progress of Research and Development. Proceedings of the Fourth

Asian Apicultural Association International Conference, Kathmandu, Nepal, 23-28 Mar

1998, pp. 178-182.

Partap, U., and Verma, L.R. 1993. Pollination of vegetable crops using the Asian hive bee, Apis cerana, in the Kathmandu Valley, Nepal. Pollination in Tropics. Proceedings of the International Symposium on Pollination in Tropics, 8-13 Aug 1993, Bangalore, India, pp. 184-186.

Partap, U., Shukla, A.N. and Verma, L.R. 2000. Comparative foraging behaviour of Apis cerana and Apis mellifera in pollinating peach and plum flowers in Kathmandu Valley,

Nepal. Asian Bees and Beekeeping: Progress of Research and Development. Proceedings of the Fourth Asian Apicultural Association International Conference, Kathmandu,

Nepal, 23-28 Mar 1998, pp. 193-198.

227

Partap, U.M.A., Partap, T.E.J., Yonghua, H.E. 2001. Pollination failure in apple crop and farmers' management strategies in Hengduan mountains, China. Pp. 225-230 In Benedek,

P., and Richards, K. W. (Eds.), Pollination: Integrator of Crops and Native Plant Systems.

Proceedings of the Eight International Pollination Symposium, Mosonmagyaróvár,

Hungary, 10-14 Jul 2000. 353 pp. ISHS Acta Horticulturae Book No. 561.

Patil, B.S., and Viraktamath, S. 2000. Pollinator fauna of sesamum in Dharwad. Insect

Environment 5 (4): 171-172.

Patil, B.S., and Viraktamath, S. 2001. Foraging behaviour of two species of honeybees on sesamum. Karnataka Journal of Agricultural Sciences 14 (3): 796-798.

Patil, B.S., and Viraktamath, S. 2005. Influence of bee attractants on pollinator visitation and yield parameters in sesamum. Pp. 195-204 In Raju, A. J. S. (Ed.), Changing Trends in Pollen Spore Research, Today and Tomorrow's Printers and Publishers, New Delhi.

Paull, R.E., and Duarte, O. 2011. Litchi and longan. pp. 221-251. In Paull, R. E. and

Duarte, O. (Eds.). Tropical Fruits, CABI, Wallingford, UK

Paydas, S., Etý, S., Sevýnc, S., Yasa, E., Derýn, K., Kaska, N. and Kaftanoglu, O. 2000.

Effects of different pollinators to the yield and quality of strawberries. Pp. 209-215 In

Proceedings of the XXV International Horticultural Congress. Brussels, Belgium, 2-7

Aug 1998. ISHS Acta Horticulturae Book No. 522.

228

Pham, H.D., Pham T.K, Nguyen, D.V., unpublished. Su dung cac giong ong mat thu phan cho dua chuot trong nha kinh o Hanoi nam 2006 (Pollination of cucumber by honeybees in the greenhouse in Vietnam, 2006).

Phipps, R. 2012. International honey market update [online]. Available from http://us1.campaign- archive1.com/?u=5fd2b1aa990e63193af2a573d&id=0329ead12b&e=67d930bdae

[accessed 15 Mar 2012].

Phoon, A.C.G. 1985. Pollination and fruit production of carambola, Averrhoa carambola, in Malaysia. Proceedings of the Third International Conference on Apiculture in Tropical

Climates, Nairobi, Kenya, 5-9 Nov 1984, pp. 129-133.

Phung, C.H. 2012. Ky thuat Nuoi ong Noi dia Apis cerana o Viet Nam (Beekeeping

Management Techniques with Apis cerana in Vietnam). Hanoi Publishing House,

Vietnam.

Phung, C.H., and Vu, L.V. 1999. Ky thuat Nuoi ong Noi dia Apis cerana o Viet Nam

(Beekeeping Management Techniques with Apis cerana in Vietnam). Agricultural

Publishing House, Hanoi, Vietnam.

Phuong, H. and Anh, T. 2012. Tang gia tri xuat khau thanh long (Improve value of dragon fruits for export). [online]. Available from http://truyenthongkhoahoc.vn/vn/Tang- gia-tri-xuat-khau-thanh-long-c1045/Tang-gia-tri-xuat-khau-thanh-long-n2959 [accessed

29Mar 2012].

229

Piña, H. H., Montaña, C. and Mandujano, M.d.C. 2007. Fruit abortion in the Chihuahuan- desert endemic cactus Opuntia microdasys. Plant Ecology. 193 (2): 305-313.

Plant Protection Research Institute. 2005. Ky thuat trong nhan. Trong Ky thuat Trong va

Tham canh Xoai, Nhan, Vai (Longan cultivation technique). pp. 32-44 In Lan, L. (Ed.),

Cultivation and Management Techniques for Mango, Longan, and Lychee. Agricultural

Publishing House, Hanoi, Vietnam.

Pleasants, J.M., and Wendel, J.F. 2010. Reproductive and pollination biology of the endemic Hawaiian cotton, Gossypium tomentosum (Malvaceae). Pacific Science 64 (1):

45-55

Popenoe, W. 1917. The pollination of the mango. U.S. Department Agricultural Bulletin

542. 20pp, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants.

United States Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Prabucki, J., Orlowski, M. and Chuda-Mickiewicz, B. 1982. The effects of pollinators on the yield and quality of cabbage seeds. Zapylanie Roslin Warzywnych. III Seminarium,

28 II - 1 III 1979, pp. 200-215. (Only the abstract seen).

Prasad, R., Chand, H. and Singh, R. 2000. Effect of honeybee, Apis mellifera pollination on the yield and germination of onion, Allium cepa seeds. Shashpa 7 (1): 53-55.

Prescott-Allen, R., and Prescott-Allen, C. 1990. How many plants feed the world?

Conservation Biology 4: 365-374.

230

Priess, J.A., Mimler, M., Klein, A.M., Schwarze, S., Tscharntke, T. and Steffan-

Dewenter, I. 2007. Linking deforestation scenarios to pollination services and economic returns in coffee agroforestry systems. Ecological Applications 17 (2): 407-417.

Primack, R. B. and Silander, J. A. 1975. Measuring the relative importance of different pollinators to plants. Nature 255 (5504): 143-144.

Rajagopal, D., and Eswarappa, G. 2005. Pollination potentiality of honeybees in increasing productivity of guava in Karnataka. Pp. 131-141 In Raju A.J.S (Ed.),

Changing Trends in Pollen Spore Research, Today and Tomorrow's Printers and

Publishers, New Delhi, India.

Ramalho, M., Kleinert-Giovannini, A. and Imperatriz-Fonseca, V.L. 1990. Important bee plants for stingless bees (Melipona and Trigonini) and Africanized honeybees (Apis mellifera) in neotropical habitats: A review. Apidologie 21: 469–488, cited by Marco

Júnior, P.D., and Coelho, F.M. 2004. Services performed by the ecosystem: Forest remnants influence agricultural cultures' pollination and production. Biodiversity

Conservation 13 (7): 1245-1255.

Ramesh, R.G., Kuberappa, G.C. and Shwetha, B.V. 2007. Quantitative and qualitative improvement of ridge gourd (Luffa acutangula) under different pollination modes. Indian

Bee Journal 69 (1): 66-70.

Rashad, S.E., Ewies, M.A. and El-Rabie, H.G. 1979. Pollinators of peanut (Arachis hypogaea L.) and the effect of honeybees on its yield. Proceedings of the Fourth

231

International Symposium on Pollination, Maryland, USA, 1978, pp. 227-230. (Only the abstract seen).

Rathi, A., and Sihag, R.C. 1993. Foraging modes and foraging rates of different bee pollinators of pigeon pea (Cajanus cajan (L.) Millionsp). Pp. 93-95 In Pollination in the

Tropics. Proceedings of the International Symposium on Pollination in Tropics, 8-13 Aug

1993, Bangalore, India.

Raw, A., and Free, J.B. 1977. The pollination of coffee (Coffea arabica) by honeybees.

Tropical Agriculture 54 (4): 365-370.

Reddi, E.U.B. 2000. Fly-pollination in the cashew nut tree, Anacardium occidentale L.

(Anacardiaceae). Journal of Palynology 35/36: 107-115.

Reddi, E.U.B. 2003. Pollinator problems of cashew farmer in India. A Review. Ecology,

Environment and Conservation 9 (3): 295-296.

Reich, L. 1991. Uncommon fruits worthy of attention. A Gardenner’s Guide. Addition-

Wesley, Reading, MA cited by William, H. 2002. Notes on economic plants. Economic

Botany 56 (2): 198-206.

Reuters. 2009. Vietnam sees coffee export volume steady in 2009-2010 [online].

Available from http://www.intellasia.net/news/articles/agriculture/111262371.shtml

[accessed 27 Apr 2009].

Rhodes, J. 2002. Cotton pollination by honeybees. Australian Journal of Experimental

Agriculture 42 (4): 513-518.

232

Ricketts, T. 2004. Tropical forest fragments enhance pollinator activity in nearby coffee crops. Conservation Biology 18 (5): 1262-1271.

Ricketts, T.H., Regetz, J., Steffan-Dewenter, I., Cunningham, S.A., Kremen, C.,

Bogdanski, A., Gemmillion-Herren, B., Greenleaf, S.S., Klein, A.M., Mayfield, M.M.,

Morandin, L.A., Ochieng, A., Potts, S.G. and Viana, B.F. 2008. Landscape effects on crop pollination services: Are there general patterns? Ecology Letters 11 (5): 499-515.

Robinson, W.S., Nowogrodzki, R. and Morse, R.A. 1989. The value of honeybees as pollinators of U.S. crops. Part 1. American Bee Journal 129 (6): 411-423.

Rodriguez, G. A., and Cervancia, C. 1999. Insect pollination of passion fruit Passiflora edulis Sims var. “flavicarpa” Degener. Philippine Journal of Science 128 (2): 119-124.

Rogers, D.J. 1965. Some botanical and ethnological considerations of Manihot esculenta.

Economic Botany 19 (4): 369-77.

Roldan, S.A. and Guerra-Sanz, J.M. 2006. Quality fruit improvement in sweet pepper culture by bumblebee pollination. Scientia Horticulturae 110 (2): 160-166.

Rose, R., Dively, G.P. and Pettis, J. 2007. Effects of Bt corn pollen on honeybees:

Emphasis on protocol development. Apidologie 38 (4): 368-377.

Roubik, D.W. 2002. The value of bees to the coffee harvest. Nature (London). 417

(6890): 708.

233

Ruszkowski, A., and Gosek, J. 1995. Food plants and the economic importance of

Hylaeus species of the subgenus Paraprosopis Popov (Hymenoptera, Colletidae).

Pszczelnicze Zeszyty Naukowe 39 (2): 255-264. (Only the abstract seen).

Ruszkowski, A., Gosek, J. and Wojdaszka, J. 1993. Food plants of Hylaeus communis.

Pszczelnicze Zeszyty Naukowe 37: 209-216. (Only the abstract seen).

Sabara, H.A., and Winston, M.L. 2003. Managing honeybees (Hymenoptera: Apidae) for greenhouse tomato pollination. Journal of Economic Entomology 96 (3): 547-554.

Sachdeva, Y., Bhatnagar, P. and Gulati, R. 2003. Relative abundance and foraging behaviour of Apis spp. on sesamum (Sesamum indicum) flowers. Annals of Plant

Protection Sciences 11 (2): 281-283.

Sadeh, A., Shmida, A. and Keasar, T. 2007. The carpenter bee Xylocopa pubescens as an agricultural pollinator in greenhouses. Apidologie 38 (6): 508-517.

Saeed, S., Sajjad, A., OhSeok, K. and YongJung, K. 2008. Fidelity of Hymenoptera and

Diptera pollinators in onion (Allium cepa L.) pollination. Entomological Research 38 (4):

276-280.

Sahu, S.K. 2011. Localized food systems: The way towards sustainable livelihoods and ecological security - a review. Journal of Animal and Plant Sciences 21 (2): 388-395.

Sajjad, A., Saeed, S. and Masood, A. 2008. Pollinator community of onion (Allium cepa

L.) and its role in crop reproductive success. Pakistan Journal of Zoology 40 (6): 451-456.

234

Sakagami S.F., Ohgushi R., and Roubik D.W. 1990. Natural History of Social Wasps and

Bees in Equatorial Sumatra. Hokkaido University Press, Sapporo, Japan.

Salakpetch, S., Chandraparnik, S. and Hiranpradit, H. 1992. Pollen grains and pollination in durian, Durio zibethinus Murr. Acta Horticulturae 321: 636-640, cited by Klein et al.

2007. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society of London. Series B, Biological Sciences 274 (1608): 303-313.

Sanchez Junior, J. L. B., and Malerbo-Souza, D. 2004. Pollinators frequency on cotton flowers and production. Acta Scientiarum - Agronomy 26 (4): 461-465.

Santos, S.A.B. 2004. Pollination of cucumber (Cucumis sativus) by stingless bees

(Hymenoptera, Meliponini). Proceeding of the 8th IBRA International Conference.

Tropical Bees and VI Encontro sobre Abelhas, p. 689, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie. 37 (2): 293-315.

Santos, S.A.B., Bego L.R., and Roselino A.C. 2004. Pollination in tomatoes,

Lycopersicon esculentum, by Melipona quadrifasciata anthidioides and Apis mellifera

(Hymenoptera, Apinae), Proceeding of the 8th IBRA. International Conference. Tropical

Bees and VI Encontro sobre, Abelhas, p. 688, cited by Slaa et al. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie. 37 (2): 293-315.

Sapir, G., Goldway, M., Shafir, S. and Stern, R.A. 2007. Multiple introduction of honeybee colonies increases cross-pollination, fruit-set and yield of “Black diamond”

Japanese plum (Prunus salicina Lindl.). Journal of Horticultural Science and

Biotechnology 82 (4): 590-596.

235

Sarfaraz, I. 2011. Do longan trees need cross-pollination from another tree? [online].

Available from http://www.ehow.com/info_10036528_longan-trees-need- crosspollination-another-longan-tree.html [assessed 30 Apr 2012].

Sarker, A.M. 2004. Effect of honeybee pollination on the yield of rapeseed, mustard and sesame. Geobios (Jodhpur) 31 (1): 49-51.

Sarto, M.C.L., Peruquetti, R.C., and Campos, L.A.O. 2005. Evaluation of the neotropical stingless bee Melipona quadrifasciata (Hymenoptera: Apidae) as pollinator of greenhouse tomatoes. Journal of Economic Entomology (98): 260–266.

Sasaki, K., Takebayashi, T. and Utsunomiya, N. 1998. Fruit set and fruit growth of

“Irwin” mango as affected by honeybee pollination. Japanese Journal of Tropical

Agriculture 42 (3): 159-162. (Only the abstract seen).

Schittenhelm, S., Gladis, T. and Rao, V.R. 1997. Efficiency of various insects in germplasm regeneration of carrot, onion and turnip rape accessions. Plant Breeding 116

(4): 369-375.

Sekhar, M.R., Singh, S.P., Mehra, R.B. and Govil, J.N. 2001. Effect of barrier crop on yield out-crossing and pollinator activity in pigeon pea (Cajanus cajan L.). Indian Journal of Genetics and Plant Breeding 61 (2): 125-128.

Sharma, S.K., Raj, T. and Joshi, A.K. 2008. Quality of carrot seeds influenced by insect pollinators under field and controlled conditions. Seed Research 36 (1): 109-111.

236

Shelar, D.G., and Suryanarayana, M.C. 1981. Preliminary studies on pollination of coriander (Coriandrum sativum L.). Indian Bee Journal 43 (4): 110-111.

Shipp, J.L., Whitfield, G.H. and Papadopoulos, A.P. 1994. Effectiveness of the bumblebee, Bombus impatiens (Hymenoptera, Apidae), as a pollinator of greenhouse sweet pepper. Scientia Horticulturae 57 (1-2): 29-39.

Shishikin, E.A. 1946. Honeybees in the service of cotton pollination. Pchelovodstvo. 23 (5-

6): 31-32, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants.

United States Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Sihag, R.C. 1988. Effect of pesticides and bee pollination on seed yield of some crops in

India. Journal of Apicultural Research 27 (1): 49-54.

Sihag, R.C., and Rathi, A. 1994. Diversity, abundance, foraging behaviour and pollinating efficiency of different bees visiting pigeon pea (Cajanus cajan (L.)

Millionsp.) blossoms. Indian Bee Journal 56 (3): 187-201.

Singh, D., Bakhshi, J.C., and Singh, K. 1970. Flowering and fruiting behaviour of ber cultivar Banarsi Karaka. Punjab Horticultural Journal 10 (1 & 2): 21-28, cited by

Dhaliwal, J.S., and Bal, J.S. 1998. Floral and pollen studies in ber (Ziziphus mauritiana

Lamk.). Journal of Research Punjab Agricultural University 35 (1): 36-40.

Singh, H.B., Kapila, R.K., Sharma, J.K. and Lata, S. 1996. Effect of insect pollination on seed yield and its quality in turnip (Brassica rapa L.). Seed Research 24 (1): 48-50.

237

Singh, M.P. 1984. Studies on the activity of some insect pollinators on jujube (Ziziphus mauritiana Lamk.). Entomology 9 (3): 177-180. (Only the abstract seen).

Singh, M.P., Singh, K.I. and Devi, C.S. 2000. Foraging behaviour of Apis cerana himalaya on sunflower and rape seed. Asian Bees and Beekeeping: Progress of Research and Development. Proceedings of the Fourth Asian Apicultural Association International

Conference, Kathmandu, Nepal, 23-28 Mar 1998, pp. 199-202.

Singh, R., and Garg, R. 1994. Relative abundance of insect visitors and pollination requirements of citrus. Pest Management and Economic Zoology 2 (2): 143-146.

Singh, R.P., and Singh, P.N. 1992. Impact of bee pollination on seed yield, carbohydrate- composition and lipid-composition of mustard seed. Journal of Apicultural Research 31

(3-4): 128-133.

Sinha, S.N., and Chakrabarti, A.K. 1993. Insect pollination in carrot seed crop. Seed

Research 20 (1): 37-40.

Siqueira, K.M.M., L.H.P., Kiill, L.H.P., Martins, C.F., Lemos, I.B., Monteiro, S.P. and

Feitoza, E.D.A. 2008. Comparative study of pollination of Mangifera indica L. in conventional and organic crops in the region of the Submedio São Francisco Valley.

Revista Brazileira De Fruticultura 30 (2): 303-310. (Only the abstract seen).

Skowronek, J., Jabloński, B. and Szklanowska, K. 1985. Effects of insect pollinators on fruit-setting of 6 varieties of strawberry. Pszczelnicze Zeszyty Naukowe 29: 205-229.

(Only the abstract seen).

238

Skrebtsov, M.F. 1964. The problem of abundant pollination of cotton by honeybees.

Trud. Nauch. Issled Inst. Pchelov. 246-264, cited by McGregor, S.E. 1976. Insect

Pollination of Cultivated Crop Plants. United States Department of Agriculture.

Agriculture Handbook. No. 496. Washington D.C.

Slaa, E.J., Sánchez Chaves, L.A., Malagodi-Braga, K.S. and Hofstede, F.E. 2006. Stingless bees in applied pollination: Practice and perspectives. Apidologie 37 (2): 293-315.

Somnuk, B., and Suavansri, T. 2005. Study on pollinating cultivars for fruit setting of

“Khom” litchi. Pp. 111-116 In Proceedings of the Second International Symposium on

Litchi, Longan, Rambutan and other Sapindaceae Plants, Chiang Mai, Thailand, 25-28

Aug 2003,.

Spangler, H.G., and Moffett, J.O. 1979. Pollination of melons in greenhouses. Gleanings in Bee Culture. 107 (1): 17-18.

Stefanidou, M., Athanaselis, S. and Koutselinis, A. 2003. The toxicology of honeybee poisoning. Veterinary and Human Toxicology 45 (5): 261-265.

Stern, R.A., and Gazit, S. 1996. Lichi pollination by the honeybee. Journal of the

American Society for Horticultural Science 121 (1): 152-157.

Sturrock, D. 1944. Notes on the mango. Sturart Daily News, Inc., Sturart, Fla., 122 pp, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States

Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

239

Subhadrabandhu, S. 1990. Lychee and Longan Cultivation in Thailand. Rumthai

Publication, Bangkok, Thailand, 40 pp, cited by Davenport, T.L., and Stern, R.A. 2005.

Flowering. In Menzel, C.M. and Waite, G.K. (Eds.), Litchi and Longan: Botany,

Production and Uses. CABI, Wallingford, UK, pp. 87-113.

Sundararaju, D. 2000. Foraging behaviour of pollinators on cashew. Cashew 14 (4): 17-20.

Sykes, S.R. 2008. The effect on citrus fruit of excluding pollinating insects at flowering and implications for breeding new seedless cultivars. Journal of Horticultural Science and

Biotechnology 83 (6): 713-718.

Szabó, Z., Nyéki, J., Benedek, P., Erdős, Z. and Soltész, M. 2000. Bee pollination and association of apricot varieties. International Journal of Horticultural Science 6 (4): 20-24.

Tanda, A.S. 1984. Bee pollination increases yield of 2 interplanted varieties of Asiatic cotton (Gossypium arboreum L.). American Bee Journal 124 (7): 539-540.

Tel-Zur, N., and Schneider, B. 2009. Floral biology of Ziziphus mauritiana

(Rhamnaceae). Sexual Plant Reproduction 22 (2): 73-85.

Tepedino, V.J. 1983. An open-field test of Megachile rotundata as a potential pollinator in hybrid carrot seed fields. Journal of Apicultural Research 22 (1): 64-68.

Tepedino, V.J. 1997. A comparison of the alfalfa leafcutting bee (Megachile rotundata) and the honeybee (Apis mellifera) as pollinators for hybrid carrot seed in field cages.

Proceedings of the 7th International Symposium on Pollination, Lethbridge, Alberta,

Canada, 23-28 Jun 1996, pp. 457-461.

240

Tepedino, V.J., Alston, D.G., Bradley, B.A., Toler, T.R. and Griswold, T.L. 2007.

Orchard pollination in Capitol Reef National Park, Utah, USA. Honeybees or native bees? Biodiversity Conservation 16 (11): 3083-3094.

Tindall, H.D. 1994. Sapindaceous fruits: botany and horticulture. Hoticultural Reviews.

16: 143-195, cited by XuMing, H., Subhadrabandhu, S., Mitra, S.K., Ben-Arie, R. and

Stern, R.A. 2005. Origin, history, production and processing. 2005. pp. 1-23. In Menzel,

C. M. and Waite, G. K. (Eds.), Litchi and Longan: Botany, Production and Uses. CABI,

Wallingford, UK.

Ting, L., ShuPing, L., Min, M., Yan, M. and Jiang, L. 2010. Flowering biological characteristics of Ziziphus jujuba Million. Acta Botanica Boreali-Occidentalia Sinica 30

(8): 1589-1594.

Toit, A.P. 1994. Pollination of avocados, mangoes and litchis. Inligtings Bulletin -

Instituut Vir Tropiese En Subtropiese Gewasse 262: 7-8.

Toit, A.P., and Swart, D.J. 1994. Pollination of mango in the Letsitele Valley during the

1993 flowering season: Second report. Yearbook - South African Mango Growers'

Association 14: 67-68.

Tra, T. 2012. Nghien cuu nang nang suat giong nhan chin muon (Study on improving production of late harvest longan) [online]. Available from http://www.vietnamplus.vn/Home/Nghien-cuu-nang-nang-suat-giong-nhan-chin- muon/20125/138781.vnplus. [accessed 15 Mar 2012].

241

Tran, T.T. 2004. Hoi Dap ve Nhan Vai (Question and Answer on Longan and Lychee).

Agricultural Publishing House, Hanoi, Vietnam.

Uji, T. 1987. Pollination in Nephelium lappaceum var.”Lappaceum”. Berita Biologi 3:

31-34. (Only the abstract seen).

Vaissière, B.E. 1991. Honeybee stocking rate, pollinator visitation, and pollination effectiveness in upland cotton grown for hybrid seed production. Pp. 359-363 In

Proceedings of the Sixth International Symposium on Pollination, Tilburg, Netherlands,

27-31 Aug 1990,.

Vaissière, B.E., Moffett, J.O. and Loper, G.M. 1984. Honeybees as pollinators for hybrid cotton seed production on the Texas high plains. Agronomy Journal 76: 1005-1010.

Valiente-Banuet, A., Gally, R.S., Arizmendi, M.C. and Casas, A. 2007. Pollination biology of the hemiepiphytic cactus Hylocereus undatus in the Tehuacan Valley, Mexico.

Journal of Arid Environments 68 (1): 1-8.

Van Deynze, A.E., Sundstrom, F.J. and Bradford, K.J. 2005. Pollen-mediated gene flow in California cotton depends on pollinator activity. Crop Science 45 (4): 1565-1570.

Vasilevskiĭ, V.A., Timin, N.I. and Pyzh'yanova, L.G. 1994. Wild bee Osmia for pollination of carrot. Kartofel' i Ovoshchi (5): 23-24. (Only the abstract seen).

Veddeler D., Olschewski R., Tscharntke T., Klein A.M. 2008. The contribution of non- managed social bees to coffee production: new economic insights based on farm-scale yield data. Agroforestry Systems 73 (2): 109-114.

242

Velthuis, H.H.W., and van Doorn, A. 2006. A century of advances in bumblebee domestication and the economic and environmental aspects of its commercialization for pollination. Apidologie 37 (4): 421-451.

Verdieva, M.G., and Ismailova, M.K. 1960. The influence of bee pollination on the increase of the crop from feed squash. Pchelovodstvo 37 (9): 40-41, cited by McGregor,

S.E. 1976. Insect Pollination of Cultivated Crop Plants. United States Department of

Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Verma, L.R., and Partap, U. 1994. Foraging behavior of Apis cerana on cauliflower and cabbage and its impact on seed production. Journal of Apicultural Research 33 (4): 231-236.

Vichitrananda, S., and Somsri, S. 2008. Tropical fruit production in Thailand. Pp. 33-46

In Proceeding of the International Workshop on Tropical and Subtropical Fruits, Chiang

Mai, Thailand, 15-18 January 2007. ISHS Acta Horticulturae Book No 787.

VietNamNet. 2009. Cashew exports facing decline [online]. Available from http://www.lookatvietnam.com/2009/04/cashew-exports-facing-decline.html [accessed

27 Apr 2009].

Viraktamath, S., and Nachappa, M.S. 2004. Relative abundance of insect pollinators on

Bt and non-Bt cotton hybrids at Dharwad. Insect Environment 10 (4): 166-168.

Vishweshwaraiah, C.H., Eswarappa, G., Kuberappa, G.C., Sreenatha, T.N. and

Sreeramulu, K.R. 2002. Foraging activity and pollinating efficiency of different species

243

of honeybees on quantitative and qualitative parameters of guava (Psidium guajava L.).

Pest Management and Economic Zoology 10 (2): 179-182.

Vithanage, V. 1990. The role of the European honeybee (Apis mellifera L.) in avocado pollination. Journal of Horticultural Science 65 (1): 81-86.

VOV-SGGP. 2006. Để cây bơ trở thành cây hàng hoá (To get avocado to be commodity)

[online]. Available from http://www.qdnd.vn/qdnd/qdndcuoituan.kinhte-xahoi.1777.qdnd

[accessed 27 Oct 2008].

VTC. 2008. MARD: Food, Food stuff demand to rise 40 per cent as Tet nears [online].

Available from http://vietnamtradepoint.com/index.php?option=com_contentandtask=viewandid=244and

Itemid=11 [accessed 25 Nov 2008].

Vu, H.C. 2000. Trong cay An qua o Viet Nam (Cultivation of Fruit Trees in Vietnam).

Agricultural Publishing House, Hanoi, Vietnam.

Wadlow, R.V. 1970. Pollination of crops in Florida. Pp. 61-63 In The Indispensable

Pollinators, Ark. Agr. Ext. Serv. Misc. Pub. 127, cited by McGregor, S.E. 1976. Insect

Pollination of Cultivated Crop Plants. United States Department of Agriculture.

Agriculture Handbook. No. 496. Washington D.C.

Wafa, A.K. and Ibrahim, S.H. 1960. The effect of the honeybee as a pollinating agent on the yield of clover and cotton. Cairo University, Faculty of Agricultural Bulletin. 206,

244

44p, cited by McGregor, S.E. 1976. Insect Pollination of Cultivated Crop Plants. United

States Department of Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Waite, G.K., and Hwang, J.S. 2002. Pests of litchi and longan. Pp. 331-359 In Peña, J. E.,

Sharp, J. L., and Wysoki, M. (Eds.), Tropical Fruit Pests and Pollinators: Biology,

Economic Importance, Natural Enemies and Control, CABI, Wallingford, UK.

Waller, G.D., Moffett, J.O., Loper, G.M. and Martin, J.H. 1985a. An evaluation of honeybee foraging activity and pollination efficacy for male-sterile cotton. Crop Science

25: 211-214.

Waller, G.D., Vaissiere, B.E., Moffett, J.O. and Martin, J.H. 1985b. Comparison of carpenter bees (Xylocopa varipuncta Patton) (Hymenoptera, Anthophoridae) and honeybees (Apis mellifera L.) (Hymenoptera, Apidae) as pollinators of male-sterile cotton in cages. Journal of Economic Entomology 78 (3): 558-561.

Wani, I.A., Bhat, M.Y., Lone, A.A. and Mir, M.Y. 2010. Unfruitfulness in fruit crops: causes and remedies. African Journal of Agricultural Research 5 (25): 3581-3589.

Wanichkul, K., and Noppapun, U. 2009. Floral biology and fruit development of Ziziphus mauritiana Lamk. 'Bomple'. Pp 312-318 In Proceedings of the 47th Kasetsart University

Annual Conference, Kasetsart, 17-20 March, 2009. Kasetsart University, Bangkok,

Thailand. (Only the abstract seen).

245

Warakomska, Z., Kolasa, Z. and Wróblewska, A. 1985. The biology of flowering and pollination of umbelliferous vegetables. Part I. Dill (Anethum graveolens L.). Acta

Agrobotanica 35 (1): 69-78.

Weiss, J., Nerd, A. and Mizrahi, Y. 1994. Flowering behavior and pollination requirements in climbing cacti with fruit crop potential. HortScience 29 (12): 1487-1492.

WenBin, Z. 1999. Analysis of the causes of unusual flower and fruitlet dropping of longan in 1999 and its correction. South China Fruits 28 (6): 31-32.

Westerkamp, C., and Gottsberger, G. 2000. Diversity pays in crop pollination. Crop

Science 40 (5): 1209-1222.

Wilkaniec, Z., and Maciejewska, M. 1996. Utilization of honeybee (Apis mellifera L.) for the pollination of the greenhouse cultivated strawberry. Pszczelnicze Zeszyty Naukowe

40 (2): 227-234.

Wille, A., Orozco, E. and Raabe, C. 1983. Pollination of Sechium edule in Costa Rica.

Revista De Biología Tropical 31 (1): 145-154. (Only the abstract seen).

William, H. 2002. Notes on economic plants. Economic Botany 56 (2): 198-206.

Williams, P. 1987. Pollination of melons. Australian Bee Journal 68 (7): 18-21.

Willmer, P.G., and Stone, G.N. 1989. Incidence of entomophilous pollination of lowland coffee (Coffea canephora); the role of leafcutter bees in Papua New Guinea. Entomologia

Experimentalis et Applicata 50 (2): 113-124.

246

Wilson, R.L., Widrlechner, M.P. and Reitsma, K.R. 1991. Pollination methods for maintaining carrot germplasm collections. Plant Genetic Resources Newsletter 85: 1-3.

Witter, S., and Blochtein, B. 2003. Effect of pollination by bees and other insects on the production of onion seeds. Pesquisa Agropecuária Brazileira 38 (12): 1399-1407. (Only the abstract seen).

Wojtowski, F., Wilkaniec, Z. and Szymas, B. 1980. Efficiency of some insect pollinators on onion and carrot seed crops. Roczniki Akademii Rolniczej W Poznaniu 120: 153-159.

(Only the abstract seen).

Wojtowski, F., Wilkaniec, Z. and Szymas, B. 1982. Preliminary results of research on insect pollination of carrot and onion seed crops. Zapylanie Roslin Warzywnych. III

Seminarium, 28 II - 1 III 1979, pp. 294-314. (Only the abstract seen).

Wolfenbarger, D.O. 1962. Honeybees increase squash yields. Fla. Agr. Expt. Sta.

Sunshine State Agr. Res. Rpt. 7 (1): 15, 19, cited by McGregor, S.E. 1976. Insect

Pollination of Cultivated Crop Plants. United States Department of Agriculture.

Agriculture Handbook. No. 496. Washington D.C.

Wong, K.C. 2000. Longan production in Asia. Food and Agriculture Organization of the

United Nations.[online]. Available from http://www.fao.org/docrep/003/x6908e/x6908e08.htm#6.%20BOTANICAL%20DESCRI

PTION [assessed 29 Apr 2012].

247

Wongsiri, S., and Chen, P. 1995. Effects of agricultural development on honeybees in

Thailand. Bee World 76 (1): 3-5.

Wu, S.H. 2010. Production status of longan in China. Pp. 37-44 In Proceedings of the III

International Symposium on Longan, Lychee, and other Fruit Trees in Sapindaceae

Family, Fuzhou, China, May 2010. ISHS Acta Horticulturae Book No. 863.

Wulfrath, A., and Speck, J.J. [n.d]. La Flora Melifera. Enciclopedia Apicola, Folleto 28, ed. 2, 96 pp. Ediciones Mexicanas, Mexico, D. F. (In Spanish), cited by McGregor, S.E.

1976. Insect Pollination of Cultivated Crop Plants. United States Department of

Agriculture. Agriculture Handbook. No. 496. Washington D.C.

Wysoki, M., Berg, M.A., Ish-Am, G., Gazit, S., Peña, J.E. and Waite, G.K. 2002. Pests and pollinators of avocado. Pp. 223-293 In Peña, J. E., Sharp, J. L., and Wysoki, M.

(Eds.), Tropical Fruit Pests and Pollinators: Biology, Economic Importance, Natural

Enemies and Control, CABI, Wallingford, UK.

XuMing, H., Subhadrabandhu, S., Mitra, S.K., Ben-Arie, R. and Stern, R.A. 2005.

Origin, history, production and processing. 2005. pp. 1-23. In Menzel, C. M. and Waite,

G. K. (Eds.), Litchi and Longan: Botany, Production and Uses. CABI, Wallingford, UK.

Yao Chun, C. 1993. Apiculture in China, Agricultural Publishing House, Beijing, China.

Young Duck, C., Man Young, L. and Young Il, M. 2001. Pollination on strawberry in the vinyl house by Apis mellifera L. and A. cerana Fab. Pp. 257-262 In Pollination:

248

Integrator of Crops and Native Plant Systems. Proceedings of the Eight International

Pollination Symposium, Mosonmagyaróvár, Hungary, 10-14 Jul 2000,

Yuan, D.Y., Wang, S., Gu, Z.Y., Duan, J.H. and Li, F.D. 2009. Blossom and fruit drop and anatomic observations on embryonic development of Ziziphus jujuba Million. cv.

“Hunanjidanzao”. Pp. 347-356 In Proceedings of the International Jujube Symposium,

Baoding, China, 21-25 Sept 2008. ISHS Acta Horticulturae Book No. 840.

Yücel, B., and Duman, I. 2005. Effects of foraging activity of honeybees (Apis mellifera

L.) on onion (Allium cepa) seed production and quality. Pakistan Journal of Biological

Sciences 8 (1): 123-126.

Zaitoun, S.T., Al-Ghzawi, A., Shannag, H.K., Rahman, A. and Al-Tawaha, M. 2006.

Comparative study on the pollination of strawberry by bumblebees and honeybees under plastic house conditions in Jordan Valley. Journal of Food, Agriculture and Environment

4 (2): 237-240.

Zietsman, P.C., and Botha, F.C. 1992. Flowering of Ziziphus mucronata subsp. mucronata (Rhamnaceae): Anthesis, pollination and protein synthesis. Botanical Bulletin of Academia Sinica 33 (1): 33-42.

249

APPENDIX

Jujube cultivars

“Dao Muon” F8 F12

Odds Odds Odds Time Intervals P 95% CI P 95% CI P 95% CI ratio ratio ratio

0.7128 – 1.5106 – 1.0537 – 09:00 – 07:00h 1.50 0.911 3.24 0.0001 2.18 0.0215 3.1646 6.9444 4.5249

0.7199 – 0.5571 – 0.6386 – 11:00 – 09:00h 1.39 0.9594 1.03 1 1.19 1 2.6667 1.9120 2.2173

0.5316 – 0.4288 – 0.3979 – 13:00 – 11:00h 0.99 1 0.79 0.9975 0.74 0.9576 1.8315 1.4577 1.3587

0.6313 – 0.5583 – 0.2513 – 15:00 – 13:00h 1.14 1 1.02 1 0.50 0.0586 2.0576 1.8797 1.0101

0.0204 – 0.0494 – 0.0485 – 18:00 – 15:00h 0.07 0.0001 0.15 0.0001 0.17 0.0002 0.2312 0.4310 0.6035

250

Appendix 2 Odds ratios and their confidence intervals for comparisons of the interaction between time of day and floral branch positions for proportions of branches with insect visitors present during scans. (Odds ratio is the ratio of numbers of visitors at one time interval divided by the numbers of visitors at the preceding time interval; P = statistical probability; 95% CI = 95% confidence interval).

Floral branch positions

High outer Low outer Low inner

Time Time Intervals Times P 95% CI Times P 95% CI P 95% CI s

1.3316 – 1.0395 – 0.6553 – 09:00 – 07:00h 2.31 0.0001 1.98 0.0246 2.32 0.6665 4.0161 3.7594 8.1967

0.4072 – 0.5757 – 1.0460 – 11:00 – 09:00h 0.71 0.8136 1.03 1 2.31 0.0254 1.2484 1.8587 5.0761

0.5086 – 0.5285 – 0.3274 – 13:00 – 11:00h 0.89 1 0.94 1 0.68 0.9429 1.5625 1.6750 1.4265

0.4000 – 0.4098 – 0.5241 – 15:00 – 13:00h 0.70 0.7251 0.74 0.9386 1.14 1 1.2210 1.3228 2.4876

0.0839 – 0.1012 – 0.0036 – 18:00 – 15:00h 0.20 0.0001 0.25 0.0001 0.03 0.0001 0.4907 0.6270 0.3084

251