Studies on plant characteristic diversity, flowering, pollination and fruit bagging in pitaya (hylocereus spp )

國立屏東科技大學農園生產系 博士學位論文 紅龍果性狀多樣性、開花、授粉與果實套袋之研究 Studies on plant characteristic diversity, flowering, pollination and fruit bagging in pitaya Hylocereus spp.. ABSTRACT Student ID: P10111005

國立屏東科技大學農園生產系

博士學位論文

紅龍果性狀多樣性、開花、授粉與果實套袋之研究

Studies on plant characteristic diversity, flowering, pollination

and fruit bagging in pitaya (Hylocereus spp.)

指導教授:顏昌瑞博士 (Chung-Ruey Yen, Ph D.) 陳幼光博士 (Yu-Kuang Chen, Ph D.)

研究生: 陳丁河 (Tran Dinh Ha)

中華民國 104 年 5 月 5 日

研究生：陳丁河 指導教授：顏昌瑞博士、陳幼光博士

論文摘要內容：

本研究之目的在調查火龍果之：(1)形態變異及果實品質、(2) 30 種基因型之開花物候性及其開花對夜間光照處理之反應、(3)授粉需求、以及(4)常見品種及優良選系之適合套袋材質等。本研究自 2013 年 4 月至

2015 年 4 月連續在國立屏東科技大學進行 2 年。依 IUPOV 標準檢定標準，總計調查了 35 種火龍果植株之莖、花及果實等形態性狀。結果顯示供試

之 30 種不同基因型之火龍果植株，其莖及果實性狀可歸類於 4 種不同之群別。第一群為具有白或淡粉紅色果肉，莖直線狀且具有波浪狀突起的

莖脊(如 H undatus 或其相關種)。第二群為紅色果肉，莖直線狀，且莖脊 平滑(如 H polyrhizus 或其相關種)。第三群為深紅色果肉，莖彎曲且莖脊 凹陷(如 Hylocereus sp.)。第四種為深紅色肉，莖直線狀且莖脊有波浪狀 突起(如 Hylocereus sp.)。此外，研究發現有 9 種基因型在一些重要的果

實特性，如果實大小、果重、果皮厚度、果肉率、與糖度等育種目標上具有市場發展的潛力。

在屏東正常火龍果約自 5-6 月至 10 月間開花，在高溫下開花期則會早些。不同基因型之植株開花波數及每株每季之開花數也不同，分別

約在 3-6 波及 9-40 朵花之間。在供試的 30 種不同基因型植株中，有 3 種白色及 3 種深紅色果肉之基因型具有完全自交親和性 (F-SC)，有 2 種深

紅色果肉基因型為部分自交親和性(P-SC)，其餘 22 種紅肉或深紅色果肉則為完全自交不親和性。在冬季利用 100 瓦白熾燈於晚間 10 時至翌日上

午 2 時連續照光 4 小時之人工夜間光照處理來測定植株之開花敏感度。相較於白肉種，紅及深色紅果肉種需要較低溫度進行花芽誘導。越早進行暗期中斷(10 月 10 日開始) 成功率越高。所有的基因型或品種，與夏天光照處理者比較，冬季處理之植株其開花波數及開花數較少，但冬果甜度較高，且部分基因型還有果實較大的現象。

研究授粉方法、花粉源對 4 種的火龍果(越南白、潮洲 5 號、

不同之營養系植株之著果率 (FSPs) 及果實鮮重 (FFWs) 有不同之影響。’越南白’在人工或天然授粉下皆有較高之著果率及鮮重。以蜜蜂到訪情形言在無天然授粉源的情況下，P-SC 型之’潮洲 5 號’或 C-SI 型之‘Orejona’

來授粉始能有適當之著果率及鮮果重，而’越南白’則在自花授粉有最佳效果。不論是否有天然授粉昆蟲的存在，在花中之花藥與柱頭的相對位置可作為是否需要人工授粉的指標。

套袋對 3 種火龍果營養系 (‘越南白’, ‘竹崎劉’及潮州 5 號)果實特性及保護之影響的研究。花後 7 天即進行白紙袋 (P-WB)、黑網袋(NS-BB)、黑塑膠袋(PP-BB)、白塑膠袋(PP-WB)套袋及無套袋等 5 種處理，直到果實採收為止。結果顯示套袋會影響果實外觀、果皮厚度、果實硬度及果實受傷害程度。其中白紙袋、黑塑膠袋為最佳的處理，它能改善果實顏色及有效降低裂果等生理損傷、鳥害及污斑等。本研究之結果可應用於各品種及改進的栽培方法上，以提高火龍果之產量及品質。

關鍵詞：火龍果、植株形態、開花物候性、光照催花、自交親和性、授

粉、果實套袋

ABSTRACT

Student ID: P10111005

Title of dissertation: Studies on plant characteristic diversity, flowering,

pollination and fruit bagging in pitaya (Hylocereus spp.)

Total page: 145 pages

Name of Institute: Department of Plant Industry, National Pingtung

University of Science and Technology Graduate date: May 5th, 2015 Degree Conferred: Doctor of Philosophy Name of student: Tran Dinh Ha Advisor: Chung-Ruey Yen, Ph D

Yu-Kuang Chen, Ph D

The content of abstract in this dissertation:

The aims of these studies are to investigate (1) morphological diversity and fruit quality; (2) flowering phenology, flowering response to night-breaking treatment and breeding system of a collection of 30 pitaya genotypes; (3) pollination requirements; and (4) fruit bagging materials suitable for some typical and promising pitaya genotypes The experiments were conducted in two consecutive years (April, 2013 – April, 2015) at National Pingtung University of Science and Technology (NPUST) In total, 35 morphological traits of stem, flower, and fruit of pitaya materials were examinedbased on the Standard Test Guidelines of IUPOV (TG/271/1, 2011) Results showed a wide morphological trait variation among 30 different genotypes Based on the stem and fruit characteristics, the pitaya germplasmscould be grouped into 4 groups: Group 1 that has white or light pink flesh, and straight rid-segment with

convex rid-margin (such as the H undatus or its relatives); Group 2 that has red flesh and straight rid-segment with flat rid-margin (such as the H polyrhizus or its relatives); Group 3 that has magenta flesh, sinuous rid-

segment and concave rid-margin (such as Hylocereussp.); and Group 4 that

has magenta flesh, straight rid-segment with convex rid-margin (such as the

Hylocereus sp.) Furthermore, some important fruit traits such as fruit

size/weight, peel thickness, proportion of fruit flesh, and sweetness in breeding targets were found in 9 genotypes that may become the most promising materials of desirable fruits for markets

The natural flowering season of pitaya in Pingtung, Taiwan usually started from May-June to October and earlier flowering may occur under higher temperature conditions.The number of flowering flushes and total flowers/season/plant highly varied among genotypes with 3 to 6 waves and 9 to

40 flowers, respectively Among 30 genotypes tested, 3 white and 3 magenta flesh genotypes showed full self-compatibility (F-SC) and two magenta flesh genotypes exhibited partial self-compatibility (P-SC) whereas 22 genotypes with red or magenta flesh were completely self-incompatible (C-SI) The artificial night lighting treatment using 100 watt incandescent light bulbs to light the plants for fourcontinuous hours from 10.00 pm to 2.00 am the next day was used to test flowering sensitivity in the winter season Red or magenta flesh pitaya species required lower temperatures for flowering initiation than white flesh types Earlier night-breaking application (starting from October 10) was more successful than one month later In comparison with the summer crop season, numbers of flowering flushes and flowers induced by lighting treatment were fewer, but winter fruits were sweeter in all fruiting genotypes and bigger in several genotypes

Effects of pollination method and pollen source on fruit set/fruit characteristics and some flowering characteristics among 4 typical pitaya genotypes including ‘Vietnam White’ (‘VN-White’), ‘Chaozhou 5’, ‘Orejona’, and ‘F11’were elucidated The pollination methods differently affected fruit set percentages (FSPs) and fruit fresh weights (FFWs) among four genotypes F-

SC genotype, ‘VN-White’ obtained high FSP and FFW after hand self-, or open-pollination Due to lack of a natural pollination efficiency as honey bee

visitations, P-SC type, ‘Chaozhou 5’ or C-SI type, ‘Orejona’ and ‘F11’ had high FSPs and FFWs by only hand-cross pollination Pollen sources also affected FSPs and FFWs ‘Chaozhou 5’, ‘Orejona’, and ‘F11’ required crossing with their compatible pollen source to ensure optimal FSPs and FFWs, while ‘VN-White’ obtained the best results by selfing (its own pollen) The relative location between the anthers and the stigma in the flower may be used as an indicator of whether hand pollination is required for a pitaya cultivar grown under the conditions with or without an availability of naturally correlative pollinator(s)

The effect of bagging on fruit characteristics and physical fruit protection

in three pitaya genotypes (‘VN-White’, ‘Chuchi Liu’ and ‘Chaozhou 5’) was also studied in the summer season of 2013 Four types of bags, including paper-white bag (P-WB), net screen-black bag (NS-BB), polyethylene plastic-black bag (PP-BB), polyethylene plastic-white bag (PP-WB) bag and non-bagged (control) were applied to fruits at 7 days after anthesis and continued until harvest Fruit bagging can affect fruit appearance, peel thickness, fruit firmness and physiological fruit damage Bagging fruits with P-WB or PP-WB,

as the best treatments, could improve the fruit color and effectively reduce the loss of damaged fruits caused by physiological factors such as cracking, birds, and blemishes in the three genotypes The results obtained from this research can be applied for the improvement of varietal and cultural practices to increase the yield and quality of pitaya

Keywords: Pitaya, plant morphology, flowering phonology, lighting-flowering

induction, self-compatible, pollination, fruit bagging

ACKNOWEDGEMENTS

Firstly, I wish to express my sincere gratitude to National Pingtung University of Science and Technology (NPUST), Department of Plant Industry and Fruit crop Lab for providing me scholarship and the best conditions to study in here

Especially, I would like to give my heartfelt and profound gratitude to

my advisors, Prof Chung-Ruey Yen and Prof Yu-Kuang Chen for their devoted guidance, valuable knowledge, and providing me facilities to do my Ph.D research program I also keep in my mind of sincere thankfulness to my advisors, who always encouraged and kindly supported me during studying and living in Taiwan I am grateful to respectable and intellectual professors

of my first, qualified and defended Ph.D evaluation committees: Dr Tzu-Bin Huang, Department of Horticulture and Biotechnology, Chinese Culture University; Dr Jau-Chang Shih, Agricultural Research Institute; Dr Chih-Ping Chao, Taiwan Banana Research Institute; Dr Chu-Ying Chiou, Kaohsiung District Agricultural Research and Extension Stationfor their best valuable comments and suggestions

I would like to extend my gratitude to Dr Charles M Papa for revision

of my paper manuscripts, the instructors and the personnel of Department of Plant Industry, Department of Tropical Agriculture and International Cooperation, Office of International Affair for their academic lessons and administration, my colleagues in Fruit Crops Lab, Plant genetics Lab, Post-harvest Lab and Plant Breeding Lab for technical assistances

Finally, I would like to thank my parents, my wife and daughters, who brought to me a positive confidence, promotion, spirit and inspiration when I was studying and living far from my family

Taiwan, May 5, 2015

Tran Dinh Ha

TABLE OF CONTENTS

摘要……… I ABSTRACT……… III ACKNOWEDGEMENTS……… VI TABLE OF CONTENTS……… VII LIST OF TABLES……….XII LIST OF FIGURES……… XIV

INTRODUCTION……….1

LITERATURE REVIEW……… 5

1 Taxonomy, diversity and breeding in climbing cacti……… 5

1.1 Taxonomy and diversity in vine cacti……….5

1.2 Pitaya cultivar and breeding……… 9

2 Flowering biology and techniques of flowering induction……….14

2.1 Flowering biology of pitaya……….14

2.2 Techniques of flowering induction in pitaya………17

3 Pollination requirement for pitaya……… 20

4 Effects of fruit bagging………23

4.1 Physiological factors influenced by fruit bagging……… 23

4.1.1 Fruit size and weight……… 23

4.1.2 Fruit colour development………25

4.1.3 Fruit maturity……… 27

4.1.4 Fruit appearance and disorders ………27

4.2 Biotic factors influenced by fruit bagging………28

4.2.1 Pest control……….28

4.2.2 Disease control……… 29

4.2.3 Bird damage……… ……….29

4.3 Fruit internal quality influenced by bagging……… 30

4.3.1 Fruit firmness……… 30

4.3.2 Eating quality of fruit………31

4.4 Effect of bagging material on fruit……… 31

Chapter 1 MORPHOLOGICAL DIVERSITY AND FRUIT QUALITY IN A PITAYA GERMPLASM COLLECTION… ………33

Abstract……… 33

1.1 Introduction……… 34

1.2 Materials and methods……… 35

1.2.1 Plant materials……… 35

1.2.2 Parameters and methods………37

1.2.2.1 Stem morphology characteristics………37

1.2.2.2 Flower morphology characteristics………38

1.2.2.3 Fruit morphology characteristics………40

1.2.2.4 Fruit quality characteristics………42

1.2.3 Statistical analysis……… 42

1.3 Results and discussion……… 42

1.3.1 Morphological characteristics of the stem……… 42

1.3.2 Morphological characteristics of the flower……… 47

1.3.3 Morphological characteristics of the fruit……… 51

1.3.4 Fruit quality characteristics………56

Chapter 2 FLOWERING PHENOLOGY, MATING SYSTEMS AND FLOWERING RESPONSE TO THE LIGHTING ADDITION OF A PITAYA GERMPLASM COLLECTION.……….60

Abstract……… 60

2.1 Introduction……… 61

2.2 Materials and methods……… 63

1.2.1 Plant materials and study site……… ……… 63

2.2.2 Parameters and methods……….65

2.2.2.1 Flowering phenology……… 65

2.2.2.2 Mating systems……….65

2.2.2.3 Flowering response to lighting treatment……… 66

2.2.3 Statistical analysis……… 68

2.3 Results and discussion……… 68

2.3.1 Flowering phenology of the pitaya genotypes………68

2.3.2 Mating systems of the pitaya genotypes……… 70

2.3.3 Flowering and fruit induction by lighting treatment………… 72

2.3.4 Fruit characteristics in off-season……… 77

Chapter 3 EFFECTS OF POLLINATION METHOD AND POLLEN SOURCE ON FRUIT SET AND FRUIT TRAITS IN PITAYA………80

Abstract……… 80

3.1 Introduction……… 81

3.2 Materials and methods……… 82

3.2.1 Plant material and experimental design……… 82

3.2.2 Treatments and data collection……… 84

3.2.2.1 Flower characteristics……….84

3.2.2.2 Pollination methods and pollen sources ……….84

3.2.3 Statistical design and data analysis……… 86

3.3 Results and discussion……… 86

3.3.1 Flowering characteristics……… 86

3.3.2 Effects of pollination method……… 91

3.3.3 Effects of pollen source……… 95

Chapter 4 EFFECTS OF BAGGING ON FRUIT CHARACTERISTICS AND PHYSICAL FRUIT PROTECTION IN PITAYA………99

Abstract……… 99

4.1 Introduction………100

4.2 Materials and methods………102

4.2.1 Plant materials and experimental treatments……… 102

4.2.2 Parameter measurements……… 103

4.2.3 Statistical analysis……….104

4.3 Results and discussion ……… 104

4.3.1 Colour parameters………104

4.3.2 Fruit firmness………107

4.3.3 Fruit characteristics……… 109

4.3.4 Damaged and defective fruits……… 111

Chapter 5 CONCLUSIONS AND SUGGESTIONS……… 113

5.1 Morphological diversity and fruit quality……… 113

5.2 Flowering phenology, mating system and flowering response to the lighting addition………113

5.3 Effects of pollination method and pollen source on fruit set and fruit characteristics………114

5.4 Effects of bagging on fruit characteristics and physical fruit protection……… 115

5.5 Overall conclusions and suggestions…….……….115

REFERENCES……… 117

Appendix Table 1 The monthly mean day-length, monthly average, maximum and minimum temperatures from Oct 2012 – Mar 2015, during the experimental period in Pingtung, Taiwan……….………… 140

Appendix Table 2 Publication credits of Ph D program in Plant Industry Department………141

BIOGRAPHICAL SKETCH OF AUTHOUR……… 142

LIST OF TABLES

Table 1 The name, species and origin of 30 pitaya genotypes used for the

plant characteristics study………36

Table 2 Morphological characteristics of the stem in 30 pitaya genotypes grown in Pingtung, Taiwan………43-44 Table 3 Morphological characteristics of the flower in 30 pitaya genotypes grown in Pingtung, Taiwan………48-49

Table 4 Morphological characteristics of the fruit in 30 pitaya genotypes grown in Pingtung, Taiwan………53-54

Table 5 Some fruit quality traits of 30 pitaya genotypes grown in Pingtung, Taiwan………58-59 Table 6 The name, estimated species, flesh colour and origin of 30 pitaya genotypes used in this study……… 64

Table 7 Flowering season, number of flowering cycles and flowers/plant/ season, floral and fruiting stages of 30 pitaya genotypes under the summer

2014 in Pingtung……….69 Table 8 Matting systems and fruit set, fruit weight resulted by involved pollination types of 30 pitaya genotypes……….71

Table 9 Flowering sensitivity, flowering and fruiting induction by lighting treatment of 30 pitaya genotypes in two off-seasons (Oct 10, 2013 – Mar 10,

2014 and Nov 10, 2014 – Mar 31, 2015)……….73-74 Table 10 Comparison of fruit traits between fruits produced during the summer

2013 and those formed during the winter 2013 - 2014 in 16 pitaya genotypes 78

Table 11 Descriptions of the four pitaya genotypes used for the pollination study……… 83

Table 12 Distance between the anthers and stigma, pollen viability and germination at two times of day in four pitaya genotypes grown in Pingtung 88

Table 13 Effect of pollination method on the fruit set percentage in four pitaya genotypes grown in Pingtung……… 92

Table 14 Effect of pollination method on fruit characteristics in four pitaya genotypes grown in Pingtung……… 93

Table 15 Effect of separate pollen source on fruit set percentage in four pitaya genotypes grown in Pingtung……… 95Table 16 Effect of separate pollen source on fruit fresh weights (g) in four pitaya genotypes grown in Pingtung……… 96

Table 17 Effect of separate pollen source on TSS content (oBrix) in four pitaya genotypes grown in Pingtung……… 98

Table 18 Effects of fruit bagging on skin colorimetric parameters in three pitaya genotypes………105 Table 19 Effects of fruit bagging treatments on fruit characteristics in three pitaya genotypes………109 Table 20 Effects of fruit bagging on physically damaged and defective fruits (%) in three pitaya genotypes………112

LIST OF FIGURES

Figure 1 Dendrogram constructed according to UPGMA cluster analysis, based

on the similarity index of Nei and Li (1979), showing the genetic relationships

among Hylocereus and Selenicereus species……….8

Figure 2 Fruits of three diploid Hylocereuss pp., H megalanthus, two

homoploid and two interploid hybrids ……….11

Figure 3 Fruit morphology of the parental lines and of the SC hybrids……… 12 Figure 4 The three main dragon fruit cultivars and breeding in Vietnam… 14

Figure 5 Flowering morphology and stages of pitaya (Hylocereus spp.)… 15 Figure 6 The cross section of pitaya (Hylocereuss pp.) flower at anthesis…16

Figure 7 Pitaya flowers pollinated by honey bees (Panel A) or a bat (Panel B).17 Figure 8 The lighting procedure applied to off-season production in Vietnam 19

Figure 9 Fruit morphology a, Fruit from maternal diploid H monacanthus accession 89-028; b, Fruit from paternal diploid H monacanthus accession 97-403; c–d, Fruits from autotetraploid line D-88 from combinanation of H monacanthus 89-028 and 97-403; c, Following pollination with a mix of pollen from diploid Hylocereus spp donors; d, Following self-pollination 21 Figure 10 Fruit morphology a, Fruit from parental tetraploid H megalanthus

accession 90-002; b, Aborted fruits from different autooctaploidlines…… 22

Figure 11 The collection of 30 pitaya genotypes preserved at NPUST orchard……….35 Figure 12 The stem morphology of 30 pitaya genotypes preserved in Pingtung, Taiwan……….45

Figure 13 The mature bud morphology of 30 pitaya genotypes grown in Pingtung, Taiwan……….50

Figure 14 Fruit morphology of 30 pitaya genotypes grown in Pingtung, Taiwan……….55

Figure 15 Lighting treatment applied to test flowering response of 30 pitaya gentotypes in off-season at NPUST……….… 67

Figure 16 Winter fruit appearance of 6 genotypes producing a higher average fruit weight than those produced in summer fruits……… 79

Figure 17 Fruit morphology in four pitaya genotypes (Hylocereus spp.) used

to study pollination in Taiwan ‘VN-White’ (Panel A), ‘Chaozhou 5’ (Panel B), ‘Orejona’ (Panel C) and ‘F11’ (Panel D) 83

Figure 18 The morphology of fully opening flower in four pitaya genotypes

‘VN-White’ (Panel A), ‘Chaozhou 5’ (Panel B), ‘Orejona’ (Panel C) and ‘F11’ (Panel D) 87

Figure 19 Honey bee visitations to pitaya flowers in the early morning (Panel A), and in the late morning (Panel B) in the NPUST orchard……….89

Figure 20 Pollen viability observed under a microscope after 1-hour incubation in four pitaya genotypes ‘VN-White’ (Panel A), ‘Chaozhou 5’ (Panel B), ‘Orejona’ (Panel C) and ‘F11’ (Panel D)………… 90 Figure 21 Pollen germination observed under a microscope after 12-hour incubation in four pitaya genotypes ‘VN-White’ (Panel A), ‘Chaozhou 5’ (Panel B), ‘Orejona’ (Panel C) and ‘F11’ (Panel D) 91 Figure 22 Comparison in fruit size produced by hand cross- and open-polliantion in ‘Orejona’ (Panel A), ‘F11’(Panel B)……… ………94

Figure 23 The high FFWs in four genotypes by crossing with their compatible pollen ‘VN-White’ pollinated by its self-pollen (Panel A),

‘Chaozhou 5’ pollinated by ‘VN-White’ pollen (Panel B), ‘Orejona’ pollinated by ‘F11’ pollen (Panel C) and ‘F11’ pollinated by ‘Orejona’ pollen (Panel D)……… 97

Figure 24 Five treatments, P-WB (Panel A), NS-BB (Panel B), PP-BB (Panel C), PP-WB (Panel D) and non-bagged (Panel E) used to study pitaya fruit bagging……… 102

Figure 25 Fruit appearance bagged with P-WB, NS-BB, PP-BB, PP-WB and without bagging in three pitaya genotypes are presented in the 1st, 2nd, 3rd, 4th and 5th columns, respectively……… 106 Figure 26 Effects of fruit bagging treatments on fruit firmness in three pitaya genotypes……… 108

INTRODUCTION

Dragon fruits or pitayas, night-blooming vine cacti, of the genus

Hylocereus (Berger) Britton and Rose arenative to Mexico, central and South

America with a wide range of habitat ecosystems, including coastal areas, high mountains and tropical rain forests (Hunt, 2006) They have a high economic potential as exotic fruit cropsin arid regions due to their possession

of a Crassulacean acid metabolism (CAM) pathway-exceptional drought tolerance (Raveh et al., 1998; Nerd et al., 2002b; Nobel and De La Barrera,

2004; Mizrahiet al., 2007)

Along with ornamental objects, their fruits are considered as one of favourite vegetable-fruit food types in the market Fruit flesh is mildly sweet, juicy, delicate aroma, and provide a resource of usefully nutritive value: rich

in fiber, vitamin, vital minerals, protein and antioxidants, and the seeds are high in polyunsaturated fatty acids (Le Bellec et al., 2006) In addition to being eaten draw, edible fruit is fresh vegetable material for processing-industrial food products as juice, jellies, yogurts, marmalades, jams, wine, and beverages (Wybraniec and Mizrahi, 2002; Chuah et al., 2008; Zainoldin and Baba, 2012) Due to be rich in antioxidants, its help to reduce the incidence of degenerative diseases such as arthritis, arteriosclerosis, cancer, heart disease, inflammation, brain dysfunction, diabetes and neutralize toxic substances such as heavy metal (Khalili et al., 2006; Halimoon and Hasan, 2010)

Pitaya agronomic practices are easy and less expensive due to low maintenance cost One of the biggest advantages of this crop is its precocious yielding ability Once planted it grows for about 20 years, and one hectare could accommodate up to 800 plants Dragon fruit is a fast return fruit crop with production in the second year after planting and full production within five years with regular bearing (Barthlott and Hunt, 1993; Mizrahi and Nerd,

1999, Zee et al., 2004; Le Bellec et al., 2006)

Due to rising demand of vine cactus fruit production over the last several years, it has been increasingly cultivated on commercial scale and

currently is being marketed worldwide Some 20 countries, including Mexico, Colombia, Ecuador, the United Stated, Vietnam, Thailand, Malaysia, Sri Lanka, Philippines, Hawaii, Israel, Okinawa, Southern China and Northern Australia currently grow this fruit crop (Nobel and De La Barrera, 2004)

The taxonomy of pitaya species are classified by distinctly morphology

as stem, flower and fruit characteristics A mong vine cati of genera

Hylocereus, four species H undatus, H monacanthus (syn.H polyrhizus), H costaricensis and Selenicereus megalanthus (syn H megalanthus) and their

hybrids are being cultivated on commercial scale (Tel-Zur et al., 2004b; Hunt, 2006) Fruit size and characteristics in these genera determine their economic

value Hylocereus (undatus, monacanthus and costaricensis) bear large fruits (200 - 600 g) H undatus is characterized by fruit oblong, peel light-red with large scales; pulp white while H monacanthus ischaracterized by fruit oblong,

peel dark-red with large scales, pulp violet-red (scarlet, purple) Nearly

similar to H monacanthus, H costaricensis species shows fruit round, peel

dark-red with large scales, pulp violet-red In contrast to three above species,

S megalanthus produces sweeter, taste fruit but lighter weight (80 – 300 g)

with form: fruit oblong, peel yellow with tubercles and spines, pulp white (Weiss et al., 1994; Ligchtenzveig et al., 2000; Tel-Zur et al., 2004a, 2011)

Pollination plays an important role in setting fruit and improving fruit

size H undatus and H megalanthus were self-fruitful (self–compatible), setting fruit with self-pollen, while H costaricensis and H monacanthus were

self-unfruitful (self–incompatible), setting fruit with cross-pollination Both

self-fruitful and self-unfruitful (Hylocereus) species produce the highest fruit

set and largest fruit obtained by interclonal or interspecific crossing (Weiss et al., 1994; Ligchtenzveig et al., 2000; Dad and Mizrahi, 2005) The pollination efficiency was related to the mating systems of species cultivars, pollination method, pollen source as well as environmental conditions Hence, understanding pollination requirements for selected cultivars growing under cultivating conditions is a requisite for orchard design and management

At growing sites outside the native region, three species H undatus, H costaricensis and H monacanthus were collected, bred and domesticated to

become commercial crops, showing adaptable growth with environmental conditions in Asian countries Over the past ten years, dragon fruit has emerged as one of potential fruit crops in Taiwan.It is grown in the whole country but Southern part isthe main area of production In 2011, the total country area harvested was about 720 hectares, producing 14,200 metric tons.Japan was the main overseas market (Freshplaza, 2013) However, Taiwan is a dragon fruit country importer from Vietnam (Hoa et al., 2006).With suitable climate and soil, pitaya production area in Taiwan is forecast to extend in the future

Dragon fruit belongs to the long day plant with natural flowering and production during warmer months (Mizrahi and Nerd., 1999; Merten, 2003).The normal bearing period of pitaya in Taiwan is from May to October Likewise some tropical countries, in Taiwan flowering of some pitaya species could be induced by breaking the dark period with supplemental lighting in off-season, allowing off-season production The fruits produced during the cool season in Taiwan brought more commercial benefits due to good fruit quality and high cost (Zee et al., 2004)

The quality of the fruit is established on the tree and influenced by the environment Bagging, a physical protection technique commonly applied to many fruits, is not only effectively used for improving the fruit quality, fruit appearance but also prevents the damages of diseases and pests, extreme environment conditions to fruits, decreasing pesticide residues, fruit drop and cracking, increasing commercial value (Kitagawa, 1992; Ni et al., 2011; Liu

et al., 2013) Bagging of individual fruit takes much labour, but the fruit industry will not exist without bagging even though the labour is very expensive It is widely believed that bagging with different material bags led

to different results in varied kinds of fruits and environmental condition of cultivation Pitaya is a non-climacteric and perishable fruit that reaches the best eating quality when harvested ripe However, until recently the official

studies and reports of fruit bagging in pitaya were still limited in Taiwan as well as in other countries

The genetic diversity in cultivated cacti was limited by the small number of progenitors and the loss of genetic variation during cultivation (Nobel, 2002) For breeding pitaya programsin Taiwan, a collection of 30

different red skin pitaya (Hylocereus spp.) genotypes from some local regions

and other countries was carried out and these genotypesare being grown at the orchard of National Pingtung University of Science and Technology (NPUST), Taiwan These materials are a valuable material sourcefor research

From horticultural and commercial points of view, the general goals of the present study are to explore the diversity of the collection of 30 pitaya genotypes in terms of plant characteristics, reproductive biology for germplasm application, and focus on pollination requirement, fruit bagging in order to improve fruit quality and yield.The specific objectives are:

1 Classifying and evaluating 30 different pitaya genotypes based on plant characteristics, flowering phenology and matting systems, flowering and fruiting induction resulted by the lighting treatment in off-season are for breeding programs and material preservation

2 Indicating pollination requirements of some typicalor promising pitaya genotypes is the aim of proposing the orchard cultivar-design and agro-managements that can improve pollination efficiency

3 Investigating the effect of fruit bagging on fruit characteristics and the bagging role in physical fruit protection is for purpose of determining suitable types of bags, which can enhance fruit quality and effectively prevent physical factors and pests damaging fruit

LITERATURE REVIEW

1 Taxonomy, diversity and breeding in climbing cacti

1.1 Taxonomy and diversity in vine cacti

Vine cacti are night blooming epiphytes, endemic to the Americas, and

belong to Cactaceae, subfamily Cactoideae, tribe Hylocereeae (Britton et

Rose) Buxbaum (Barthlott and Hunt, 1993) According to the New Cactus Lexicon (Hunt, 2006) that the taxonomy of this group was revised, the genera

Hylocereus (Berger) Britton et Rose, Selenicereus (Berger) Britton et Rose and Epiphyllum Haw.comprise 14, 12 and 12 species, respectively The distribution area of the members of the Hylocereeae tribe is mainly the forests

of Central America, with a few species extending throughout tropical

America (Barthlott and Hunt, 1993) The genus Hylocereus is native to

Central America, the West Indies, and the Northern areas of South America

Selenicereus is indigenous to the tropical Americas and the Caribbean region, and the habitat of the genus Epiphyllumis mainly found in Central America

with a few species extending to the West Indies and South America (Barthlott and Hunt, 1993) These three genera bear attractive and exotic edible fruits with numerous small soft seeds (Mizrahi and Nerd, 1999; Lichtenzveig et al., 2000)

Barthlott and Hunt (1993), and Hunt (2006) described these three

genera as follows: Hylocereus species are usually characterized by their

triangular stems/normally three-angled stems, which often produce aerial roots, and their large, globose, ovoid, or oblong fruits with broad scales

Selenicereus species are typified by stems that have 2 – 12 ribs with

irregularly giving off aerial rootsand globose, ovoid, or oblong fruits, with

spines, hairy-spines, or bristles Epiphyllum species are distinguished by terete

stems at their base that later fatten out, and ovoid or oblong fruits that are naked or, rarely, have hairy-spines or bristles

Among three cactus genera, Hylocereus species and their hybrids have

been commonly used as cultivars The classification and distinction of species and varieties in this genus is difficult due to the high intra-and interspecific hybridization (Tel-zur et al., 2004b), which has caused some taxonomic confusion Therefore, there have been several studies on morphological characterization of the genotypes In Mexico, a study evaluated the diversity

of H undatus through morphological characterization and sexual

compatibility of 6 genotypes (Castillo et al., 2005), who concluded that the reproductive characteristics were the most important to separate the genotypes

Other study conducted a morphometric analysis of 21 genotypes of H undatus using 47 characteristics in a multivariate analysis but, unlike the

previous studies, the authors found that the most important variables to separate genotypes were those of the stems (Grimaldo et al., 2007) A study conducted in Israel, a morphological and genetic characterization of 64

accessions of Hylocereus, Selenicereus and Epiphyllum was carried out,

including variables such as: nuclear DNA content, stomatal density and length and plant production parameters, and found that there is high genetic variability among accessions and therefore, excellent prospects for domestication and conservation of these genetic resources (Tel-zur et al.,

2011) In Colombia, Mejíaet al (2013) studied on In situ morphological characterization of Hylocereus spp (Family: Cactaceae) genotypes from

Antioquia and Córdoba with 23 descriptors of stems and fruits using GPS and description The stem descriptors differed in spine number, contour and margin hardness, areole distance and stem color; the first three are important for species distinction The latest study, characterization ofgenetic

relationship of dragon fruit accessions (Hylocereus spp.) by morphological

traits and ISSR markers, was firstly employed to discriminate 50 accessions recently selected in China, as well as to evaluate their genetic relatedness (Tao et al., 2014), showing that high variations in morphological traits between or/and within the wild and cultivated lines and a significant correlation between the data of two evaluation methods

Cytological observations have shown that almost species of Hylocereus and Selenicereus were diploid (2n = 2x = 22), whereas only S megalanthus was tetraploid (2n = 4x = 44) (Beard, 1937; Spencer, 1955; Lichtenzveig et al., 2000) Unlike other species of Selenicereus, S megalanthus had a three- winged stem, like that of Hylocereus, and spiny fruits like those of the Selenicereus It was classified by Britton and Rose (1963) as a separate genus, Mediocactus, thereby implying both an intermediate morphology and an

intermediate taxonomic status between the latter two genera Bauer’s

placement of this species in Hylocereus reflects the close affinity between the tetraploid taxon and the diploid Hylocereus species (Bauer, 2003) The fact that a tetraploid taxon S megalanthus shared morphological features with two diploid taxa Hylocereus spp, which were also cross compatible might imply

an allotetraploid origin by natural intergeneric hybridization between closely

related diploid taxa (Hylocereus and Selenicereus species) (Lichtenzveig et al.,

2000)

Since vine cactus species have a long juvenile phase and germplasm identification based on morphological characteristics that encountered problems in identifying the different genotypes Development of molecular techniques would be valuable for a suitable taxonomic description of these species and for future breeding and genetic studies of these genera DNA isolation from cacti is notoriously difficult because they contain high amounts

of polysaccharides and secondary metabolites which form insoluble complexes with nucleic acids during extraction (Guillemaut and Drouard, 1992) In effort to find methods for more suitable DNA extraction from these species, an effective procedure as a modification of the original CTAB method proposed by Tel-Zur et al (1999) had been successfully applied to the

root samples of epiphytic Cacti of the Genera Hylocereus and Selenicereus.

By using molecular DNA marker (RAPD), genotypes of Hylocereus and of Selenicereus species were distinguished and identified genetic

relationship among them (Tel-Zur et al., 2004a) (Figure 1) These results were consistent with the accepted taxonomic classification of the genera studied

The RAPD results support the hypothesis regarding the allopolyploid (rather

than autopolyploid) origin of H megalanthus species

Figure 1 Dendrogram constructed according to UPGMA cluster analysis, based on the similarity index of Nei and Li (1979), showing the genetic

relationships among Hylocereus and Selenicereus species

With crossing and analyzing putative hybrids by chromosome counts and morphological traits, confirming the ploidy level by fluorescent in situ hybridization (FISH) of rDNA sites, identifying the putative diploid genome donors by genomic in situ hybridization (GISH), this indicated that overall

sequence composition of H megalanthus is similar to that of H ocamponis and S grandiflorus High sequence similarity was also found between the parental genomes of H monacanthus and H megalanthus in one triploid hybrid and the most successful cross was H monacanthus × H megalanthus, suggesting that H megalanthus is closer to H monacanthus than to H

undatus or Hylocereus sp Therefore, Hylocereus and Selenicereus close

relationship is supported by their intergeneric hybrid produced in cultivation (Tel-Zur et al., 2004b, 2005)

Cacti were threatened by the loss and degradation of their habitat and illegal collection (Oldfield, 1997; Boyle and Anderson, 2002); the most important factor causing the loss of biodiversity is intensive land use (Zak et al., 2004) The ‘‘Red List’’ of threatened species of the International Union for the Conservation of Nature (IUCN, 2008) provides a categorized list of species based on their relative risk of extinction at a global scale; the list includes at least 104 cacti species (7% of all species) as vulnerable to extinction (Mihalte and Sestras, 2012)

Currently, the vine cacti grown on a commercial scale as fruit crops

belong to four species of the genera Hylocereus (Ortiz and Carrillo, 2012)

The genetic diversity in cultivated cacti was limited by the small number of progenitors and the loss of genetic variation during cultivation (Nobel, 2002) and most of the domesticated cacti grown for fruit or ornamental flowers apparently originated from a relatively narrow germplasm base (Mihalte and Sestras, 2012)

1.2 Pitaya cultivar and breeding

According to the New Cactus Lexicon the genus Hylocereus comprises

14 species (Hunt, 2006), almost cultivated dragon fruit varieties belong to

four species: H undatus, H monacanthus (syn H polyrhizus), H costaricensis and H megalanthus (syn S megalanthus) and their hybrids (Ortiz and Carrillo, 2012) H undatus cultivar is widely cultivated in the

World Its fruit is oblong shape, white flesh and red peel with large scales (Figure 2) The exceptional points of this cultivar are large fruit size, high yield potential In addition, it was self-compatible and therefore allows growers to design single-cultivar orchards without the need for cross

pollination, reducing labour cost H monacanthus also bears large fruit with

oblong shape, violet-red pulp and dark-red peel with large scales Nearly the

same as H monacanthus, H costaricensis species produces large fruit but round shape, violet-red pulp and dark-red peel with large scales H monacanthus and H costaricensis reveal advantages of fruit size and attractive flesh appearance In addition, stems of H monacanthus are covered

with wax, which may impart an advantage under stress conditions (Cisneros and Tel-Zur, 2010) The major disadvantage of them is self-incompatible and growing outside the native region, the plants are routinely cross pollinated to guarantee good yields (Weiss et al., 1994; Nerd and Mizrahi, 1997) This procedure requires that fresh pollen be readily available, thus indicating that the orchard be designed to include at least two cultivars that bloom

simultaneously (Weiss et al., 1994) Yellow pitaya (H megalanthus) is

mainly cultivated in Columbia and Israel (Nerd and Mizrahi, 1998) It has aoblong fruit, yellow peel with tubercles and spines, white flesh (Weiss et al.,

1994; Dag and Mizrahi, 2005) Despite a desired flavor and sweet fruit, H megalanthus bear small fruit, low yields and the attendant man-power costs

required to remove the spines before marketing Because of the fruit’

excellent flavor, however, H megalanthus has been used in breeding

programs to develop elite, interspecific-interploid hybrids

The breeding programs for improving fruit quality and yields have been conducted in some countries and achieved advanced results In Israel, the practical breeding including reciprocal and interspecific homoploid,

interclonal in Hylocereus spp (2x × 2x) crosses were performed, that yielded

diploid and polyploid hybrids with improved traits As a result of their enhanced traits, some hybrids are currently being grown on a commercial

scale (Tel-Zur et al., 2004b) (Figure 2)

Figure 2 Fruits of three diploid Hylocereus spp., H megalanthus, two

homoploid and two interploid hybrids

Owing to successful tools of in vitro culture, androgenesis (Garcia et al., 2009a) and gynogenesis (Garcia et al., 2009b) formed haploid plants for producing homozygous lines, and embryo rescue (Cisneros and Tel-Zur, 2010; Cisneros et al., 2013) is used to nurture immature or weak embryos or to grow whole plants from the unviable/aborted seeds that typically result from wide

hybridizations Crosses between these Hylocereus spp and H megalanthus (4x

× 2x and 2x × 4x) yielded viable triploid, pentaploid, hexaploid and aneuploid

hybrids that have never been obtained in these crosses under natural conditions The marketability of the triploid fruits was further enhanced by their acceptable size (200 – 300 g) To achieve self-compatible (SC) hybrids that would greatly benefit growers by eliminating the need to grow complementary cultivars for

cross-pollination, two breeding techniques were applied Autopolyploidization with antimitotic reagents resulted in the break-down of the self-incompatible (SI) system in the autotetraploid lines (Cohen and Tel-Zur, 2012) Other method is crossing the previously selected triploid hybrids ‘S-75’ and ‘12-31’

as the female parents The crosses included a first back-cross, a new generation

of interploid interspecific hybrids (F1) using as the pollen donor the diploid

Hylocereus undatus, and a second generation of hybrids (F2) A a result, five hybrids, all of which were a cross between the triploid hybrid ‘S-75’ and the

diploid Hylocereus undatus, were found to be fully SC; they exhibited similar

fruit weights and similar total seed numbers following cross- or self-pollination (Tel-Zur et al., 2012) (Figure 3)

Figure 3 Fruit morphology of the parental lines and of the SC hybrids (a) Diploid

H polyrhizus, (b) Tetraploid H megalanthus, (c) Triploid hybrid ‘S-75’of cross H monacanthus × H megalanthus, (d) Diploid H undatus, (e to h) Hybrids of cross between the female hybrid ‘S-75’ and the donor pollen H undatus, (e) Hybrid ‘Z-

10’, (f) Hybrid ‘Z-11’, (g) Hybrid ‘Z-17’, (h) Hybrid ‘B-189’

Breeding and selection research in Taiwan and Vietnam have resulted

in self-fertile and productive pitaya varieties Some notable selections such as

‘Vietnam #1’ produce large, pink fruits (averaging 14 oz, ± 3 oz) with white flesh and high levels of total soluble solids (a measure of sweetness) of 13 – 19% Many selections are being evaluated from the red-fleshed fruit types

belonging to two closely related species, H polyrhizus and H costaricensis, and their hybrids with H undatus (Zee et al., 2004)

In Vietnam, there have been three main dragon fruit cultivars grown in commercial scales (Figure 4) The white pulp cultivars, namely Binh Thuan and Cho Gao as local varieties, have dominated about 90% growing area and production for over 100 years These cultivars are self-compatible and perform oblong fruit shape with large and green scales, an average fruit weight from 360 to 380 g, 14.9 0Brix, pH: 4.9 – 5.1, percentage of edible fruit:

73 – 75% (Hang et al., 2014) Long Dinh number 1‘LD No1’variety is hybrid

of ‘White Flesh’ local variety × ‘Red Flesh’genotype (introduced from Colombia) This cultivar had fruit weight 350 – 400 g, Brix 16%, red flesh, fruit firmness 0.7 kg/cm2, easy fruit set without hand pollination Long Dinh number 5 ‘LD No5’variety is hybrid of ‘LD No1’ × ‘White Flesh’ The hybrid dragon fruit wasvigorous, pink purple flesh, green scale, fruit weight

350 - 400 g, Brix 17%, fruit firmness 0.96 kg/cm2 ‘LD No5’ fruit quality is better than ‘LD No1’ in rainy season

Currently, in order to recognition and evaluation of new piatya varieties, International Union for the Protection of New Varieties of Plants (IUPOV, 2011) proposed “The Guidelines for the Conduct of Tests for Distinctness,

Uniformity and Stability (DUS)” applying to Dragon fruit (Hylocereus undatus (Haw.) Britton & Rose This document standardized and described

the fully characteristics of pitaya as stem, flower, fruit

Figure 4 The three main dragon fruit cultivars and breeding in Vietnam

2 Floweringbiology and techniques of flowering induction

2.1 Flowering biology of pitaya

Some previous studies confirmed that the natural flowering and

production season of Hylocereus spp was during the summer, from May to

October in Taiwan and Israel, from April to October in the South of Vietnam,

from April to November in Sri Lanka wheareas flowering in S megalanthus

was concentrated in autumn in Israel Flowering occurred in 3 - 7 cycles (flushes) per year depending on different species and cultivated condition As many as 4 - 8 flowering cycles may occur in tropical areas, tropical areas, however in cooler climate areas it is more normal for an average of 2 - 3 cycles

to occur Study on flowering phenology, Pushpakumara et al (2005) reported that from spherical buttons emerge to flowering (anthesis) lasted about 17 days Flowers stayed open and pollination in one night to the following morning If pollination was completed and thereafter petals begin to wilt (Figure 5)

Figure 5 Flowering morphology and stages of pitaya (Hylocereus spp.)

Observing floral structure and flowering biology of pitaya, (Figure 6), flowerswere extremely showy, hermaphrodite, bell shape, large, white or cream white in colour, very fragrant to attract natural pollinators They had a thick tube bearing several long linear green scales Stamens were cream colored, and formed a showy fringe in the center and at the apex of the thick perianth tube The hollow style was cream in colour and about 26 cm long Stigma was also creamy white in colour and divided into average of 15 lobes

A single flower of dragon fruit with numerous anthers produced a massive amount of pollen Pollen grains were sticky and uniform in size (a diameter of

70 - 80 µm) After opening, flowers produced a musk smell and about 7 ml of nectar at the bottom of the floral tube (Pushpakumara et al., 2005) Weiss et al (1994) described that during flower opening and the receptive period the hollow stigma was positioned upright whilst after pollination it turns

downward The upper part of the anthers in the Hylocereus spp flowers were

at least 2 cm below the stigma (1), whereas in S megalanthus, they were at

the same height as the stigma, touching it when the flowers closed

Figure 6 The cross section of pitaya (Hylocereus spp.) flower at anthesis

The pitaya flower is nocturnal as known as the Queen of the night or moon flower They opened rapidly, starting at around 6.30 – 7.00 pm and completing by about 10.00 am the next morning If pollination has been completed, thereafter petals will begin to wilt However, if flowers have not been pollinated during the night, they willremain open till 11.00 am the next day During flower opening and the receptive period the hollow stigma was positioned upright whilst after pollination it turns downward The flower petals closed completely by daybreak (Zee et al., 2004; Pushpakumara et al., 2005)

Pitaya flowers were adapted and attracted bats or hawk moths in the evening and honey bees in the morning to pollinate (Figure7) Pollination by natural pollinators was one important factor to improve fruit set and fruit size, especially for self-incompatible species In case of absence of natural pollinators, hand pollination was routinely carried out to guarantee good yields The blooming of pitaya flowers was affected by temperature and light intensity (Zee et al., 2004; Valiente Banuet et al., 2007)

Figure 7 Pitaya flowers pollinated by honey bees (Panel A) or a bat (Panel B)

2.2 Techniques of flowering induction in pitaya

Factors affecting flowering of pitaya include shoot age, temperature, light, and application of growth regulators (Jiang et al., 2012) Regardless of day-length, extreme temperatures in summer may restrict the flowering of pitaya (Nerd et al., 2002b) Light intensity may affect the nutritional status of shoots and thus influence the number of flowers produced, but it does not

promote off-season flowering in H undatus (Khaimov and Mizrahi, 2006;

Raveh et al., 1998) Flower thinning and application of growth regulators such

as [N-(2chloro-4-pyridinyl)-N-phenylurea] (CPPU) and gibberellic acid (GA3) could accelerate flowering or extend the flowering period and therefore result

in earlier or later harvest, but they did not reverse the seasonal flowering

phenology in H undatus (Khaimov and Mizrahi, 2006) A number of other

chemicals related to floral initiation, including sodium naphthalene acetic acid,

ethephon (Ethrel ), ethephon with urea, a commercial mixture of gibberellins (GA4 and GA7) with benzyladenine known as PerlanTM7, paclobutrazol, hydrogen cyanamide, 2-ethylene chlorohydrin, KNO3, and CaC2, have been tested butdid not promote winter flowering of pitaya (Chang, 2003; Khaimov and Mizrahi, 2006; Yen and Chang, 1997)

Dragon fruit is known to be a long-day plant, requiring longer day lengths to bear flowering (Yen and Chang, 1997; Luders, 1999), so controlling bloom initiation with artificial lighting may be a possibility, thus extending the

season In Taiwan, red pitaya (Hylocereussp.) flowers between May and

October and sprouts between November and May in Taiwan The critical daylength seemed to be ≈12 hours, and night-breaking treatment would be applicable between the September and the next March equinoxes to produce off-season crops Four hour-continuous lighting between 10.00 pm and 2.00 am effectively gave off-season production from November to April The duration

of night-breaking required for flower differentiation was longer in the cooler than in the warmer season The fruits produced during the cool season in Taiwan were more desired in the market than fruits from summer crops because the off-season fruits are larger and sweeter The recommend lighting utilizes incandescent light bulbs (100 watt) at about 4 – 5 ft spacing suspended about 6 ft above the ground (Zee et al., 2004; Su, 2005; Jiang et al., 2012)

In the southern Vietnam, lighting treatment has been most commonly applied to induce off-season flowering and fruit production on commercial scales It resulted in higher flower and fruit uniformity than PGR method producing less stable flowering and fruiting Night-breaking is normally conducted in 6 - 8 hours a night time, and continuously in 15 - 25 nights depending on the season and weather, which was sufficient for flowering (Figure 8) After stop lighting of 3 to 5 days, the flower bud would appear, after that it needed 20 – 21 days to develop flower and 3 days for flower open and fertilization and then the fruit developed within 25 to 28 days to mature Totally, it takes 75 - 80 days from lighting to harvest, depending on the time

of the year in which flowering stimulation is conducted and weather condition

in the cultivation areas (Hoa, 2008)

Figure 8 The lighting procedure applied to off-season production in Vietnam

In the northern tropical region of Thailand, the lighting supplementation from fluorescent tube in short day season potentially replaced long day

condition, resulting in the flower induction in H undatus The lighting

programs were composed of two- and four-hours after sunset, (18.00 - 20.00 and 18.00 - 22.00 pm), and two-hour night-break (22.00 - 24.00 pm) The supplemented light treatments exhibited induced flower buds within 43 - 48 days The duration from the bud emergence to fruit harvest was about 112 days The night-breaking treatment yielded 67% fruit set, maximum fruit number and fruit weight The night-break lighting was apparently most effective for flower bud induction and development (Saradhuldhat et al., 2009)

However, Khaimov and Mizrahi (2006) reported that day-length had no effect on pitaya flowering in Israel This response is different from that reported by the above authors, Yen and Chang (1997) in pitaya in Taiwan, or that described by Borchert et al (2005), who found that minor changes in day-length (30 min.) induced flowering in a number of succulents These contradictory results may be explained by differences in the temperature regimes between the tropical climates of the above-mentioned studies and the sub-tropical climate of Israel Following to the previous study on off-season production of pitaya by breaking the dark period with supplemental lighting

in Taiwan (Su, 2005), light inducing with 4 hours from 10.00 pm – 2.00 am was indicated as an effective treatment

3 Pollination requirement for pitaya

Among 4 cultivated pitayas, H undatus and H megalanthus were fruitful (self–compatible), setting fruit with self-pollen while H costaricensis and H monacanthus were unself-fruitful (self–incompatible), setting fruit with

self-cross-pollination Fruit weight was positively correlated with seed numbers and a positive relationship seems to exist between the seed data and pollen sources,

pollination methods Both self-fruitful and self-unfruitful (Hylocereus) species

produce the highest fruit set and largest fruit obtained by interclonal or interspecific crossing (Weiss et al., 1994; Ligchtenzveig et al., 2000; Dad and Mizrahi, 2005; Valiente Banuet et al., 2007) Many of the varieties were not self-compatible; therefore, the flowers needed to be cross pollinated with pollen from

a different genotype or species in order to set fruit At growing sites outside the

native region of Hylocereus spp., the plants were routinely cross pollinated by

hand to guarantee good yields if lacked of natural pollinators This procedure required that fresh pollen be readily available, thus indicating that the orchard be designed to include at least two cultivars that bloom simultaneously (Weiss et al., 1994; Nerd and Mizrahi, 1997) This adds a considerable amount to the labor

cost of growing these fruit Many varieties from Asia (predominantly H undatus)

were self-compatible, and some of these were autogamous and would set fruit without the involvement of a pollen vector (Merten, 2003) Weiss et al (1994),

Dag and Mizrahi (2005) reported that Selenicereus megalanthus, the yellow

pitaya, similar fruit set and fruit weight were obtained for both self-pollinated and cross-pollinated flowers, indicating full self-compatibility but higher fruit set than that in automatic self pollination However, there was a difference in fruit weight between two authors According to Dag and Mizrahi (2005), fruit weight were significantly heavier in hand-pollinated flowers than inbagged flowers and open pollinated flowers while the same level of fruit weightin hand cross pollination and in open pollination was reported by Weiss et al (1994) The

former explanation would be due to a high number of ovules and low pollen viability or perhaps to the transfer of insufficient amounts of pollen by bees as a result of the short overlap time between bee activity and a flower's receptivity to pollination These findings contradicted the earlier hypothesis of Weiss et al (1994) that spontaneous physical contact between anthers and stigma during flower-closing (stigma were at the same height in these flowers and touch as the flower closes), would produce levels of pollination that are as efficient as in hand-pollination Moreover, this feature would be related to the fact that pollen transfer in open pollination was achieved by bee visits and by direct transfer of pollen to the stigma, which occured via physical contact between anthers and stigma during flower closing

In an attempt to improve self-fruiting in Hylocereus species, the autopolyploidization method was induced for H monacanthus and H megalanthus (Figure 9, 10)

Figure 9 Fruit morphology a, Fruit from maternal diploid H monacanthus accession 89-028; b, Fruit from paternal diploid H monacanthus accession

97-403; c–d, Fruits from autotetraploid line D-88 from combinanation of

H.monacanthus 89-028 and 97-403; c, Following pollination with a mix of pollen from diploid Hylocereus spp donors; d, Following self-pollination

Figure 10 Fruit morphology a, Fruit from parental tetraploid H megalanthus

accession 90-002; b, Aborted fruits from different autooctaploid lines

The resulting H monacanthus autotetraploids exhibited lower fruit weight,

seed number, and pollen viability than the donor plant, but it has larger pollen

grains Although the resulting H megalanthus autooctaploids had larger pollen

grains and lower pollen viability compared with the donor plant, only aborted fruits were obtained from these lines The most valuable change observed was

the breakdown of the self-incompatibility system in the H monacanthus

autotetraploid lines Hence this study showed that inducing autopolyploidization

in Hylocereus species was not an effective method to improve commercially

important horticultural traits (Cohen and Tel-Zur, 2012)

Hand pollination was carried out easily by physically removing anthers from one flower and touching them to the stigma of another or collecting the pollen and using a brush to pollinate multiple flowers Pollen was most viable

at the time of flower opening, but hand pollination was found to be successful well in during the entire period of anthesis (within 24 hours from flowers started to open) Hand pollination could thus be carried out conveniently during the morning hours, if economically feasible (Weiss et al., 1994) Another problem that often occurs was that varieties would bloom when there were no other flowers open from which to obtain pollen To circumvent these problems Metz et al (2000) developed a protocol for the long-term storage of pollen They found that the pollen must be dried to 5% to 10% moisture content by weight and stored at subfreezing temperatures Pollen stored in this manner remained viable for at least 9 months The colder the storage

temperature the larger the resulting fruit was after hand pollination This protocol would enable growers to store pollen over the winter and pollinated the first blooms of the season thus getting an earlier and larger crop

4 Effects of fruit bagging

Fruit bagging hasbeen successfully used for 70 years old (Kitagawa et al., 1992) and considered as a physical protection technique, commonly applied to many kinds of fruits It not only improves their visual quality by promoting peel colouration and reducing the incidence of disease, insect pests, mechanical damage, sunburn of the skin, agrochemical residues on the fruit, and bird damage, but can also change the microenvironment for fruit development, having multiple effects on internal fruit quality (Fan and Mattheis, 1998; Kitagawa et al., 1992; Hofman et al., 1997; Joyce et al., 1997; Tyas et al., 1998; Amarante et al., 2002a; Xu et al., 2010) Pre-harvest bagging of fruit is commonly practiced in Australia, China, Japan and Taiwan during peach, apple, pear, grape, banana, mango and loquat cultivation in order to optimize fruit quality and increase market value Some countries such

as Argentina, Chile, Mexico do not import apples unless they are bagged (Sharma et al 2014)

There have been contradictory reports on the effects of pre-harvest fruit bagging on fruit size, maturity, skin colour, flesh mineral content, and fruit quality, all of which may be due to differences in the type of bag used, the stage of fruit development when bagged, the duration of exposure to natural light following bag removal and/or fruit- and cultivar-specific responses (Hofman et al., 1997; Fan and Mattheis, 1998; Sharma et al., 2014)

4.1 Physiological factors influenced by fruit bagging

4.1.1 Fruit size and weight

Covering fruit with a bag at a particular developmental stage may influence their growth and size Reports on the effects of fruit bagging on fruit