Productions fruitieres et horticoles des regions tropicales et mediterraneennes

••••••••••••••••••••••••••••••••••••••••••• ••• •••••••••••••••••••••••••••••••••••••••••• SSN 0248-1294 Vol. 49, n°S-6, 1994 Special Issue on Tropical Orchards ORGANISME EDITEUR : CIRAD-FLHOR.

DIRECTEUR DE LA PUBLICATION : J.L. Rastoin. CONSEILLER SCIENTIFIQUE (de ce numéro): B. Aubert.

RÉDACTEURS C. Loison-Cabot, C. Baudouin.

Secrétariat de rédaction, mise en page C. Potel. Traducteur : D. Manley.

ABONNEMENTS 1995 : Publications Elsevier 141, rue de Javel 75747 Paris cedex 15, France

Revue bimestrielle. 1995, volume 50 (6 numéros). Tarifs 1995 : France : 750 FF. EU et autres pays : 900 FF EU (avec TVA) : 919 FF Amérique du Nord, centrale et du Sud US $ 215 including air delivery.

Prix de ce numéro spéclal : 255 FF

Numéro 5-6, 1994. Dépôt légal 4e trimestre 1995. CPPAP n°62005 - ISSN n°0248-1294. La reproduction totale ou partielle des articles de Fruits est autorisée sous réserve expresse de la mention d'origine. Achevé d'imprimer: Paragraphic, 31240 L'Union (Toulouse). Some of the papers presented at the September 1993 Symposium on Tropical Orchards, within the framework qf ']ourneesCIRAD-FLHOR", are published in this special issue. To address a wide readership, the articles are published in English and French, with key words in English, French and Spanisb. Tbe English versions qf the articles (text, illustrations, tables and bibliographical references) f are published in full in the first hal qf the volume, wbile mentioning the paging of the French versions that are presented in the last balf(colour pages). Note tbat all papers presented at tbe Symposium on Tropical Orchards could not be covered in this limited 1994 special issue of FRUITS. However, on the last page there is a list qf other com­ plementary papers from the Symposium that bave already been published (or are soon to be published). Tbe reader will tbus have access to all of the information conveyed at this special meeting on tropical orchards. We hope that readers will find all information they are seeking in this document. THEEDITORS

•••••••••••••••••••••••••••••••••••••••••••••••••

Ce numero special "vergers tropicaux" permet de publier une partie des exposes qui ont ete presentes a !'occasion des journees CIRAD-FLHOR de septembre 1993. Pour que leur contenu soit accessible a tous nos lecteurs, les articles sont publies en deux langues (anglais et fran\;ais), et les mots des sont trilingues (anglais, fran\;ais et espagnol). Les versions anglaises des documents (texte, illustrations, tableaux, bibliographie) sont publiees dans leur integralite en debut de volume, leur traduction en fran\;ais (textes seuls) est repartee dans la deuxieme partie de l'ouvrage (pages de couleur). Cependant, du fait du grand nombre, et du volume, des communications qui avaient ete pre­ sentees a !'occasion de ces journees CIRAD-FLHOR .. vergers tropicaux ", toutes n'ont pu etre incluses dans ce numero special 1994 de la revue Fruits. Afin que le lecteur puisse avoir, malgre tout, une connaissance exhaustive du contenu global de la manifestation, une liste des publications (deja parues ou a para'itre), complementaires de ce numero special, est pre­ sentee a la derniere page. Nous esperons que chacun pourra trouver, a la lecture de ce document, !'information qu'il recherche. LA REDACTION FRUITS Productions fruitieres et horticoles des regions tropicales et mediterraneennes

contents

I. Economy p. 331-347 11. Cropping Techniques p. 349-360 111. Genetics p. 361-408 IV. Crop Protection p. 409-431

A list of other publications of interest is presented on the last page of this issue.

sommaire

I. Economie p. 433-444 11. Techniques culturales p. 445-454 111. Genetique p. 455-485 IV. Defense des cultures p. 487-503

Une /iste des publications complementaires de ce • numero special » est presentee a la derniere page de ce numero.

vol. 49, n°S-6, septembre-octobre-novembre-decembre 1994 ••••editorial tropical orchards: the exotic implications Hot region orchards customarily evoke exoticism, rare and festive pleasurable commodities. According to this image, the orchards are often pictured as flourishing in special enclosures where growers bring together a few bountiful producing a wide range of nutritional and appealing fruits that strike the imagination. Tbis hedonistic "18th century Rousseauistic" viewpoint is no longer relevant, despite the fact that tropical fruit consumption is booming (partly based on the traditional image), and presently higher than that of temperate fruits and even other prestigious produce. Tropical fruit consumption is not always clearly linked with daily energy needs, since the fruit is often eaten between meals, but some species are now widely consumed on a regular basis, e.g. oranges, mandarins and dates. Beyond their novel attraction, other tropical fruits (grapefruits, avocados, lemons, papayas, etc.) meet a need/or dietetic and health foods. Some otherfruits, e.g. litchis, , mangosteens and rambutans, still have an emotional and symbolic impact, thus explaining their market success. Tbe wide spectrum of vitamins and flavours offered by tropical orchards is a prime attraction for modern adventurous societies seeking original products. tropical fruit production controlled by market economy laws Tropical and exotic fruit production is affected by mass consumption patterns, irrespective of the nutrient status. Producers therefore have to come up with pertinent responses to this situation: regular supplies and quality control, careful packaging and competitive pricing. CIRAD-FLHOR scientists are involved, through their research, in promoting tropical fruit products. Consumers often underestimate how much effort is required to reinforce and stabilize this production sector: - surveyingmarket trends to forecast market demand, - perfecting horticulturaltechniques, - rational management of genetic resources, - efficient, low-input disease and pest protection. Tbis special issue of FRUITS includes some of the papers presented at the symposium on Tropical Orchards held from 30 August to 5 September 1993 in Montpellier (France). A wide range of technical and scientific information that could interest both research scientists and people involved in the tropical fruit industry is reviewed. Tbis information is presented as short updates (new cropping strategies, propagation techniques), and as clear and detailed summaries (germplasm management, hybridization and selection projects, genetic analyses, ident(fication and control of new parasitesand pests). As an introduction, main trends and consumer requirements are considered from an economic perspective. There was too much information presented at the Symposium on Tropical Orchards to be covered in a single special issue, even in a double issue. Some of the symposium papers were thus published in FRUITS vol. 49 (1994), and others are to be included in the next issue (vol. 50, 1995). We would like to express our gratitude to all people who helped with this special issue, which was aimed at promoting the development of a sector that is involved in manyrural tropical areas. BERNARD AUBERT Head of the Citrus and Orchard Fruit Programme

•••• editorial

le verger tropical : un terme hautement evocateur Le verger des regions chaudes a traditionnellement evoque l'exotisme, les produits rares, pour tout dire les aliments de Jete et de plaisir. On l'a souvent imagine sous forme d'enclos oi1 l'homme aurait rassemble quelques e:,,peces vegetates genereuses lui qffrant une gamme de fruits utiles a sa sante, propres aussi a le rejouir et a le faire rever. Cette vision " dix-huitieme siecle ", hedoniste et " rousseauiste ", n 'est plus guere de mise, mhne si l'accroissement de la consommation de fruits tropicaux, du fait meme d'une certaine reminiscence de cette image, reste aujourd'hui superieur a celui des fruits temperes, voire a celui d 'autres denrees de prestige. Sans necessairement presenter de liens directs avec le besoin energetique journalier puisqu 'ils sont souvent pris en dehors des repas, certains de ces fruits sont neannioins definitivenient entres dans nos habitudes de consommation : oranges, mandarines, dattes, etc. D'autres, outre l'attrait de la nouveaute, ressortissent egalement a un souci de dietetique et d'hygiene alimentaire : pamplemousses, avocats, citrons, papayes, etc. Pour certains enfin : le litchi, la mangue, le mangoustan ou le ramboutan, la charge emotionnelle et la symbolique predominent dans l'acte d'achat. La large palette de vitamines et d'ar6mes que le verger tropical est susceptible d 'o.ffrir, constitue, il est vrai, un atout majeur pour nos societes toujours plus avides de nouveautes et d'evasion. des productions confrontees aux lois de l'economie de marche Quel que soit leur rang dans la gamme des nutriments, les productions fruitieres tropicales au exotiques n 'echappent pas aux problemes souleves par la consommation de masse, et vis-a-vis desquels le producteur doit trouver des reponses pertinentes : regularite des approvisionnements, surveillance constante de la qualite, soins apportes au conditionnement, prix competitifs. Les chercheurs du CIRAD-FLHOR contribuent, grace a leurs travaux, a promouvoir ces produits fruitiers des tropiques. Le consommateur sous estime d'ailleurs souvent la somme d'ffforts a deployer pour conf011er et stabiliser cette filiere: - suivi de la tendance des marches pour anticiper !'evolution de la demande, - perfectionnement des techniques horticoles, - gestion rationnelle des ressources genetiques, - protection phytosanitaire efficace et econome en intrants. Ce numero special de la revue FRUITS, regroupe une pa11ie des communications du colloque Vergers Tropicaux tenu a Montpellier du 30 aout au 5 septembre 1993. JI passe en revue un ensemble d'informations techniques et scientffiques susceptibles d'interesser a la fois le monde des professionnels et celui des chercheurs. Certaines de ces informations constituent de courtes notes de mise au point (nouvelles approches culturales, techniques de propagation), d'autres se presentent sous forme de syntheses claires mais sifffisamment detaillees (gestion du germplasme, travaux d 'hybridation et de selection, etudes genetiques, identification et maftrise de nouveaux parasites et ravageurs). En introduction des considerations d 'ordre economique font etat des principales tendances, voire exigences posees par le consommateur. Le volume des informations collectees !ors de ce colloque " Vergers tropicaux " depasse malheureusement le cadre d 'un simple numero special, fut-ii double. En consequence, certaines des communications qui y ant ete presentees ant d�ja ete publiees dans le volume 49 (1994) OU /e seront prochainement dans le volume 50 (1995). Nous tenons a remercier toutes les personnes qui ant bien voulu participer a la realisation de ce numero special et contribuer ainsi a promouvoir le developpement d'une filiere au se trouvent impliquees de nombreuses zones rurales des tropiques. BERNARD AUBERT Responsable du Programme agrumes et arboribulture fruitiere I. Economy

The European Market: a Promising Tropical Fruit. 0. LOEILLET [mangoes, markets, European communities, tropical frnits, exports, imports] 332-334 [mangue, communautes europeennes, frnit tropical, expo1tation, impo1tation] [mango. mercados, comunidades europeas, frntas tropicales, exportaciones, importaciones]

Tropical Fruit in the Non-French Caribbean. Crops, Exports, Trends. G. BARBEAU [Blighia sapida, Citn,1,:,� Anacardium occidentale, Artocarpus altilis, Persea americana, 335-339 Averrhoa carambola, Malpighiaglabra, Annona muricata, Psidium guajava, indica, Spondias dulcis, Manilkara zapota, Calocarpum sapota .. .] [... frnit trees, diversification, production, trade, CaribbeanJ [. .. arbre fruitier, diversification, production, commerce, Cara·ibes] [. .. arboles frntales, diversificaci6n, producci6n, comercio, Cari be]

New Challenges for the Mandarin/Mandarin-Hybrid Industry in the Mediterranean Basin. B. AUBERT [Citrns frnits, fruit growing, mandarins, development strategy, Southeast Asia, Medite1Tanean countries] 340-343 [Agrume, culture fruitiere, mandarine, strategie de developpement, Asie du Sud-Est, pays mediterraneens] [Frutas citricas, fruticultura, mandarina, estrategia de! desarrollo, Asia sudoriental, paises mediterraneos]

Refrigerated Fruit Juices. New Outlets for World Fruit Crops. D. LOEILLET [Fruit juices, soft drinks, markets, trade, pasteurizing, quality, packaging, storage] 344-347 (Jus de frnit;boisson non alcoolisee, marche, commerce, pasteurisation, qualite, conditionnement, stock.age] (Jugo de frutas, gaseosas, mercados, comercio, pasteurizaci6n, calidad, empaquetado, almacenamiento]

Fruits, vol, 49 (5-6) / Special on Tropical Orchards •• 331 The European Mango Market: a Promising Tropical Fruit

D. LOEILLET •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR Th e market appeal and consumption of mango, the 5th-ranking fruit 26, rue Poncelet 75017 Paris produced in the world, are sharply increasing in developed France countries, especially in Europe. There are many producing countries, but imports are higher from the Americas than from Africa. Fruits, vol. 49, n°5-6 p. 332-334 (English) •••••••••••••••••••••••••••••••••••••••••••••• p. 434-435 (French)

Mango is one of the most important fruits figures, the yearly growth rate for mango marketed in the world, its appeal and production is slightly less than 2%. consumption are booming in developed Most mangos are grown in developing countries, especially in Europe. countries, with India as the undisputed Improving the relation between supply leader (10 Mt - 60% of the world out­ and demand, with respect to consump­ put); then far behind come Mexico tion periods and varieties, should (854 OOO t), Pakistan (780 OOO t) and enhance market development and, in Thailand (614 OOO t). turn, improve taste quality of export man­ gos. Asia predominates in the geographical distribution of mango production, with 80% of the world output, followed by the Asia, Americas and Africa at much lower rates the main production region of 13% and 7% respectively. Figure 1 Mangos, the 5th-ranking fruit crop world­ Main trade flow directions wide, are produced extensively in about Central and South America, fo r mangos Source: ODM CIRAO.FLHOR. 70 countries. According to the latest FAO two important export regions The global trade volume for these regions is estimated at more than 250 OOO t. Mexico, with nearly 100 OOO t, is the world leader for mango exports. The US market, its major outlet, absorbs 90% of these exports. Internationally, as shown in Figure 1, three trade flow directions have develo­ ped over the years: - South and Central American countries supply the North American market, �-,. , Europe and Japan; Japan - Asia preferentially exports to countries ;.a . .\ within its own region and to the Middle �· East; .._, -Africa markets most of its fruit on the . i European market.

332 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • ECONOMY e

the United States and Europe, 1000 tonnes the main importers 50 USA, with over 76 OOO t of imports, is by far the top world import market 40 (Appendix 1). Despite a slight slump in 1992, there has been a 30% average yearly mango import growth rate in USA 30 since 1988. Although the respective import rates for 20 Canada and Japan (Appendix 1) are only, at best, a quarter that of the EU, these two markets have been increasing stea­ 10 dily over the past few years. The EU, 2nd-ranking world importer of mangos, imported 46 OOO t in 1992. The 0 development of the EU import market 75 77 79 81 83 85 87 89 91 since 1975 is illustrated in Figure 2. year Figure 2 The EU mango import market Three countries consume more than 75% Source : National customs. Processing · OOM ClRAD-FLHOR. since 1975. of the mangos imported in Europe: UK, Germany and France (Fig. 3). The UK is by far the top European consu­ mer, because of the oriental ethnic com­ munities settled there and its privileged relationships with countries such as Pakistan and India. The UK is also one of the main driving forces behind the mango market. mango producing countries and Figure 3 their market ranking Source : National customs. Processing: 0D!v1 CJRAO-FLHOR Main European mango importers in 1992. The EU, second for market volume after USA, is the most diversified market. % Most of the 90 classified mango produ­ cing countries (63 in 1992) supply less 30 than 100 t of fruit for the European mar­ ket. Brazil, Puerto Rico and South Africa 25 are the only high volume exporters on this market. 20 Of the 26 main mango sources that sup­ ply the market (98% of yearly imports), 15 14 are from the Americas or the 1985 1986 1987 1988 1989 1990 1991 1992 Figure 4 Caribbean, 8 are African and 4 Asian. ACP market shares in the EU Source , ODM CIRAD-FLHOR since 1985. Market shares for mangos sourcing from ACP (Africa/Caribbean/Pacific) countries, bound to the EU by preferential trade agreements (Lame IV), dropped by for ACP countries, followed by a further 11 points between 1985 and 1992 (Fig. 4). decline in 1992 even though the mango There was a short-lived reflation in 1991 market was developing rapidly.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 333 • e LOEILLET

There are six essential reasons why ACP supervised development countries have chronic problems in mar­ keting their mangos on the EU market The world mango market has been deve­ (generally their only solvent market): loping steadily over the past 10 years. The EU market is the most diversified, - high competition between ACP coun­ ,v ith respect to countries of origin and tries at the time (spring) when European varieties offered to the consumer. New consumption is oriented towards seasonal markets such as Spain and Italy still have local fru it ; a high consumption potential. Some - difficulties in controlling sea transport; Asian and American producing countries - landlocked situation of some countries; are now highly competitive with ACP producing countries, as African mango - predominance of one variety (cv Amelie) exports decline. which is not yet fully appreciated by African producers will have to adopt consumers because of its green coulour; three development strategies in order stall - high competition with Caribbean, further market decline: establish quality­ Central American and South American based policies (transport, packaging), mango producers; extend their supply of mangos to the - great difficulty in breaking into British European market over a longer period, and German markets; the French market and intensify their promotional campaign represents almost half of all ACP outlets. fo r mango products. e

Appendix 1 Source : National customs. A) USA mango imports between 1988 and 1993 (tonnes) * first fo ur months Year Total Mexico Growth rate 1988 34 646 27 269 + 34 1989 52 273 43 923 + 16 1990 59 007 50 922 + 56 1991 92 122 76 402 -17 1992 76 165 68 254 1993 • 20 245 17 243

B) Canadian mango imports since 1991

Tonnes 1991 1992 Total 13 796 12 782 Mexico 8 418 8 048 USA 2 333 2 555 Brazil 425 368 Venezuela 510 327 Costa Rica 8 238 Peru 231 231 Philippines 156 187 Dominican Republic 60 112 Taiwan 117 95 Thailand 67 86

C) Japanese mango imports between 1989 and 1992

Tonnes 1989 1990 1991 1992 Total 6 013 5 445 6 925 8005 Philippines 4 664 4 212 5 768 7 177 Mexico 1 206 1 158 1 036 734 China 32 26 35 38 Thailand 78 23 20 18 Taiwan 21 6 50 14.

334 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards Tropical Fruit Trees in the Non-French Caribbean. Crops, Exports, Trends

G. BARBEAU •••••••••••••••••••••••••••••••••••••••••••••• MAE/IICA Regional Fruit Project Trinidad and Tobago Some of the many fruit trees cultivated in the non-French Caribbean show a promising market potential. Fruits, vol. 49, n°5-6 p. 335-339 (English) •••••••••••••••••••••••••••••••••••••••••••••• p. 436-439 (French)

The non-French Caribbean comprises Trinidad and Dominica. Limes, especially many different countries marked by a West Indian lime, were initially cultivated wide range of histories, sizes, peoples by essential oil exporting industries, but and languages. The agricultural export began declining before World War II crops that have long prevailed are because of pest and disease problems increasingly affected by serious interna­ After the war, some countries (e.g. Belize, tional competition. These countries are Dominica and Trinidad) switched to crop­ thus attempting to set up agricultural ping oranges and grapefruit, but else­ diversification programmes with tropical where the industry is still steadily declin­ fruit production as a key component. This ing. study examines development projects for At the outset of the 1970s, Cuba launched ligneous fruit trees, marketing outlets for a major citrus planting programme; it con­ fresh and processed products and tinues to be the focus of major attention, medium-term production prospects. and 148 OOO ha is presently cropped with citrus fruits in the country. ackee Production also started decreasing in (Blighia sapida) Trinidad during the 1970s because of the

r oil boom, but it resumed in the micl- This f uit of African origin is widely 1980s. A private firm recently planted grown in Jamaica, where it is served with 1 300 ha with citrus fruits and planned to salt fish. Almost every garden has an plant a further 200 ha in 1994. ackee . This apparently ve1y hardy species bears reel fruit, which open when Other Caribbean countries such as Belize, ripened and expose shiny black seeds. Jamaica and the Dominican Republic The fruit must be eaten at a ve1y specific recently tried to develop their citrus time, otherwise it could be toxic. Ackee industry. A 1 200 ha planting program fruit is canned by several manufacturers, was just started in western Jamaica. mainly for the local market, but small Cuba, the largest exporter of fresh citrus quantities are exported to other fruits in the Caribbean region, exported Caribbean countries. 400 to 500 OOO t/annum before 1990, mainly to socialist countries, and is sup­ citrus fruits posed to have exported 600 OOO tin 1991. (Citrus sp.) Jamaica exports fresh fruit to the EU (8 300 t in 1990, 7 400 t in 1991), includ­ The citrus fruit inclust1y has f1ourishecl for ing the famous cultivars Ortanique and much of the 20th century in several Ugli. The Dominican Republic and Caribbean countries, particularly Jamaica, Dominica export small quantities to

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 335 • e BARBEAU

Guadeloupe and Martinique. Belize, West Indies University and the Caribbean Jamaica and Trinidad have their own pro­ Agricultural Research and Development cessing industries to serve domestic and Institute (CARDI) in Po1t of Spain to export markets, but Trinidad still imports improve transpo1tation conditions and concentrates from Belize (more than fruit shelf-life; the results could help to $3 million US in 1991} increase exports. The Caribbean citrus-growing indust1y With respect to A. alth'ilis var. seminffera, has begun expanding again. However, the seeded form, only the seeds are con­ there is a serious threat of a tristeza out­ sumed, either grilled or boiled. A few break, a disease that spreads rapidly in tonnes are exported to the usual bread­ this region. Most crops are grafted on fruit markets. sour orange trees and could thus be des­ troyed. Several countries (Cuba, Jamaica, Dominican Republic and Trinidad) have taken control measures. avocado (Persea americana L.) The Dominican Republic is by far the top Caribbean avocado producer, with an cashew estimated 4 OOO ha cropped with this (A nacardium occidentale L. ) fruit, and the equivalent of 12 OOO ha in Cashew trees are found throughout the scattered trees, producing a total of Caribbean, mainly in areas with a marked 130 OOO t/annum of avocados (FAO). This dry season. Cashew is not cultivated, country recently started exporting to USA apart from a few exceptional cases (60 ha and EU, mainly cvs Semi! 34, Choquette in Trinidad). The fruit and false fru it are and Hall. The fact that these cultivars pro­ consumed locally. A cashew development duce large-sized fruit does not seem to be project was set up with Brazilian assis­ problematic. tance in southern Saint Lucia. Jamaica is Cuba is the only other Caribbean country r interested in cropping this fuit to deve­ cropping avocado. There is an estimated lop its southern inland savannas; Guyana 1 800 ha of avocado plantations in Cuba, could crop cashew in areas inhabited by but so far the entire crop is absorbed by native Imiians. The region generally the domestic market. imports commercial cashews. Cashew cropping, which is possible in areas with Fifteen years ago, Dominica tried cultivat­ low precipitation, could give rise to a ing avocados for export, but the results small-scale rural agricultural indust1y, thus were disappointing. At present, there is reducing imports and stalling rural outmi­ an estimated 168 ha cropped with avo­ gration. cado. Other countries intercrop avocados for domestic consumption. Grenada and Saint Vincent export substantial, but now breadfruit decreasing, quantities to Trinidad. (A rtoca17Jus altilis Fosb.) This species is as common as coconuts and mangos in the Caribbean region. carambola There are many different ways of con­ (A verrhoa carambola L.) suming the fruit of the seedless variety. Although there are no commercial planta­ This fruit is still quite unknown outside tions (with intercropping), some countries Guyana and Surinam, which have large export their surplus production. Saint numbers of scattered trees and a few Lucia exports an average of 900 t/annum, homogeneous orchards. There are 120 ha and Dominica, Grenada and Trinidad of orchards and the equivalent of only export a few dozen tonn s, mainly 2 450 ha of scattered trees in Guyana, and to ethnic markets in Great Britain and about 10 ha of orchards and an undeter­ Canada. Research is under way at the mined number of scattered trees in

336 •• Fruits, vo l. 49 (5-6) / Special on Tropical Orchards • ECONOMY •

Surinam. Most of the crop is lost because cessed for domestic markets. Firms intend of the long transportation distances, des­ to expand production of this crop and pite the efforts of some small firms to export the fruit pulp. develop this potentially profitable crop. A small percentage of processed carambola The only countries exporting soursop are the Grenada islands and Saint Vincent, is exported to the Caribbean region. and the entire crop grows on scattered A carambola fly [Bactrocera sp. (e.g. trees. These two islands have a notewor­ Dacus dorsalis)] has recently showed up thy advantage, they have not been affec­ in Surinam from Southeast Asia, and now ted by fruit flies, Annonaceae weevils or threatens Guyana where a survey system other fruit pests. Trinidad is the main cus­ has been set up to monitor the situation. tomer (around 600 t/year), for the fresh In Surinam, an FAQ coordinated program fruit market and making pulp, which is aims at eradicating this insect. expo1ted in small quantities to Great Britain and Holland. A closely-related species, the custard West Indian cherry apple (A . squamosa L.), is also grown in (Malpighia glabra L.) Grenada and Saint Vincent and exported to Trinidad, for the fresh fruit market The West Indian cherry tree, or Barbados only. Prospects for the expansion of this cheny tree, looks like a bushy shrub. fruit are not ve1y promising since it is Red, slightly-ribbed, cherry-sized fruit are highly perishable. produced over a large part of the year, as long as there is sufficient rainfall. The fruit is rich in vitamin C and highly appre­ ciated on certain markets, particularly in guava Japan. A few years ago, when labour (Psidium guajava L.) costs got too high in Barbados, the Guava is found in all Caribbean countries. Japanese tried unsuccessfully to find Cuba probably has the largest surface other partners in the Caribbean region, area cropped with guava, but the entire but finally turned to Vietnam. crop of this highly-esteemed fruit is In the Caribbean, the other countries pro­ absorbed by the domestic market. ducing this cherry are Puerto Rico, In the other Caribbean countries, the pro­ Guyana and Surinam. The planted surface duce is mainly derived from small area is still limited but rapidly increasing. orchards and scattered trees, and the fruit Trees derived from plantlets obtained by is often infested with many types of fruit vegetative propagation are productive in flies sp.). Cuba seems to be the first year and give satisfacto1y yields. (Anastrepha the only potentially serious candidate for Juice is made on the spot with the highly developing guava as an export crop - perishable fruit, it is sold on the domestic they have bred extremely productive market. Surinam has started exporting dwarf varieties. small quantities to Barbados, which no longer produces enough to meet its own consumption needs. According to experts from the International Trade Center (ITC, mango UNCTAD/GATI Geneva), the West Indian (Mangtfe ra indica L.) cherry has a very promising international market potential. In English-speaking Caribbean countries, cv is the main mango variety. A few other varieties such as cvs , , Ceylon and Peach are also mar­ soursop keted. (A nnona muricata L.) Most of the mango trees are scattered in Soursop is cultivated on a small scale in fields, not planted in orchards. Despite market-oriented orchards only in Guyana this, there are so many mangos produced and Surinam, where it is grown to be pro- that several countries export them - it

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 337 • e BARBEAU

is the most plentiful fruit crop along sapodilla with breadfruit. Saint Lucia exports (Ma nilkara sapota Van Royen) 400-600 t/annum, Dominica around 100 t/annum, Grenada and Trinidad a sapodilla plum few dozen tonnes. (Calocarpumsa pota Merr.) Jamaica has been using Florida vanettes Sapodilla is found everywhere but is not since the mid-1980s, and has thus been cultivated commercially. Trinidad imports able to regularly increase exports to the a few dozen tonnes from Grenada and EU (more than 1 200 t in 1991); the Saint Vincent. Everywhere else this deli­ Dominican Republic is following this cious fruit is seriously infested by fruit trend. These two countries continue to flies and does not seem to have any pro­ plant mangos on large surface areas. Most mising prospects for future development. Caribbean countries have to target the Conversely, sapodilla plum is plentiful in European market because of the phyto­ Cuba and the Dominican Republic, where sanitary restrictions applied to exports it is in high demand on the domestic mar­ towards USA, as set out in the North ket. However, substantial breeding American Free Trade Agreement (NAFTA). research still has to be done on the many local types before this fruit is ready for export. ambarella (Spondias cytherea Sonner) conclusion Within a 3-year period, Grenada increa­ Concerning perennial fruit trees, citrus, sed ambarella exports from a few tonnes mango and avocado trees predominate in (1989) to about 500 t (1991) and thus terms of planted surface area (Table 1), emerged as an exporter of this fruit on and they are the most commonly expor­ the international market. The produce is ted species. essentially aimed at New York markets, Production and export of fresh citrus fruit and London and Amsterdam markets to a and processed products should continue lesser extent. The conventional Trinidad increasing in Cuba, Belize, Jamaica, market also accounts for about 200 t. 90% Dominican Republic and Trinidad. of the fruit is exported green, to be can­ However, a potential outbreak of tristeza ned and processed in traditional is a major threat to this activity. Caribbean preparations. The produce comes from scattered, intercropped trees. Two countries are expanding their mango The Grenada Ministry of Agriculture is production, Jamaica and Dominican taking an active interest in this product Republic, with the European market as and, following the setbacks experienced their main target. in 1992 due to problems of competition Avocado cropping is only important in between exporters, is trying to improve the Dominican Republic. quality and reduce grading differences. The neighbouring island of Saint Vincent Otherwise, the good export performance also traditionally exports ambarellas to of breadfruit and the progressive elimina­ Trinidad and, following Grenada's exam­ tion of technical constraints are notewor­ ple, is now targeting the international thy. market. There is considerable development poten­ The fruit of a closely-related species, tial for three fruits, so far considered as myrobalan (S. mombin L.), called "mope" being minor products, i.e. soursop, West in Surinam, is processed by small firms Indian cherry and ambarella. Research on and widely consumed as delicious fresh the first two has been undertaken by pri­ juice in this country and Guyana. The vate companies in Guyana and Surinam. species is not cultivated, it grows in natu­ Ambarella is currently very popular in ral forest stands. Grenada and St Vincent. •

338 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • ECONOMY e

Table 1 Areas cropped with perennial tropical fruit trees in the non-French speaking Caribbean (ha).

Antigua B:trbados Belize Cuba Dominica Gren:1cb Guyana J:un:1ica Dominican Sl Kitts St Lucia SI Vincent Surinam Trinidad Rep. Tob�1go

Ackee B Citrns A 8 12 23600 144000 2020 1620 8100 7000 32 2425 6000 B 60 15000 10 680 Cashew A 2 60s B Breadfruit A 1 B 29 820 Avocado A 2 3 1800 168 57 4000 7 18 2 B 12000 2 162 15 Carambola A 120 5 B 2450 Cherry A 32 36 53 B 38 Soursop A 22 B Guava A 2 22 2000 113 20 B 26 11 Mango A 12 47 585 18000 124 10 526 300 18 25 B 133 60 30000 20 545 22 250 Ambarella B 21 20 Rambutan A 3 B Saporaceae B

A: orchards casual trees B: scattered trees more common sources

COLEACP Database - Yearly Reports. Proceedings of the First Regional Workshop on Tropical Fruit Crops, Dominica, IICA, CARDI, FAO. Production and Trade Yearbooks. UWI, February 1991. Proceedings of the Caribbean workshop on tradi- Proceedings of the Second Regional Workshop tional and potential fruit tree crop develop­ on Tropical Fruit Crops, Antigua, IICA, CARDI, ment. IICA Ministry of Agriculture of UWI, December 1991. Grenada. Grenada, November 9-14, 1980. Published by IICA San Jose, Costa Rica, Proceedings of the Regional Workshop on Fruit 1981. Diversification, Guadeloupe, IICA/CIRAD, April 26-28, 1993. Port of Spain, Trinidad Proceedings of the lnteramerican Society for and Tobago, IICA/CIRAD, 139 p. Tropical Horticulture, Vol 33, 1989. International Carambola Workshop, Guyana, • ••• September 4-6, 1989.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 339 New Challenges for the Mandarin/Mandarin-Hybrid Industry in the Mediterranean Basin

8. AUBERT •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR Citrus output in the Mediterranean Basin (1 7 Mt) is much lower than BP 5035 34032 Montpellier cedex 01, in North and South America (30 Mt), although consumer markets France in these regions are of similar size. It is predicted that by the year 2020 the population of the southern rim of the Basin will have Fruits, vol. 49, n°5-6 expanded by 78%, with a probable marked increase in the domestic p. 340-343 (English) market demand for citrus fruits in these countries. European p. 440-442 (French) consumers are increasingly attracted by new seedless easy-peeling citrus fruits. Th e Mediterranean citrus industry will thus have to be substantially restructured to address the new challenges . • • • •• • •• •• • •• • •• •• •• • •• • • • • • •• • • • • • •• • • • • •• • • •

specific trends beginning of the present era. It is thus an important centre of seconda1y genetic in Mediterranean and diversification, and several different local European citrus markets types are known to originate exclusively from this region, e.g. blood oranges, compared to those in North willow-leaf mandarin (Citrus deliciosa), and South American markets clementines, etc. This diversity could be explained by the World citrus production has increased local horticulturists' practice of selecting considerably over the past 20 years. Total natural mutations. Mediterranean consu­ output is currently some 75 Mt, with mers thus gradually came to appreciate oranges ranking first, followed by manda­ fresh citrus fruit as a delicacy, and then as rins, lemons and grapefruits (Table 1). a common and easy-to-eat product. Historically, the initial citrus-producing Nowadays, the wider Mediterranean area areas were Asia and the Mediterranean includes 20 countries with overall citrus Basin, where various types were first output of 17 Mt in 1992/1993 (Table 2), domesticated. Citrus growing has long with: been a traditional occupation in the - 58% marketed as fresh fruits on domes­ Mediterranean area; Pliny the Elder, for tic markets in the producing countries instance, described citron cropping at the (400 M consumers); citrus is the main type of fruit consumed in these countries; - 29% exported as fresh fruit to markets Table 1 in Northern Europe (west and east); these World citrus production for 1992/93 (thousand tonnes, FAO statistics). non-producing countries represent a mar­ ket of 330 M consumers (excluding Oranges 55 122 73.0% Russia); Mandarins & hybrids 8 257 11.0% - 13% only is destined for juice process­ Lemons and limes 6 624 8.8% Grapefmits 4 981 6.6% ing. Comparatively, 30 Mt of citrus fr uits are Total 74 984 produced in North and South America,

340 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • ECONOMY e

mainly oranges, to a lesser extent grape­ fruits and lemons, and ve1y few manda­ Table 2 rins. Of this output: Citrus output and population of the Mediterranean area for 1992/93 - 6% is exported as fresh fruit, (FAO statistics). - 12% is absorbed by domestic fresh fruit markets, Country Citrus Population - 81 % is used for juice processing. (thousand t) (M) This breakdown highlights that New Northern Mediterraean rim 9 536 199 Albania World consumers are as a whole more 14 3 Spain r 4 704 38 oriented towards juices than f esh fruits, France whereas the trend is reversed for 30 58 Greece 1 202 10 Mediterranean consumers. There are Italy 3 440 57 302 M consumers in the North American Portugal 138 10 market, i.e. USA and Canada, whereas ex Yougoslavia 8 23 there are 460 M consumers in the Latin Southern Mediterranean rim 7 795 208.8 American market, spread throughout Algeria 35 countries. 295 26 Cyprus 345 0.7 Egypt With an output of only 17 Mt, the 1 820 60 Gaza 165 0.8 Mediterranean area obviously cannot Israel 1 460 5 meet the demand of its 730 M consumers Jordan 111 4 Lebanon (Mediterranean countries, Europe, exclud­ 450 3 ing Russia). Comparatively, the 30 Mt Libya 104 5 citrus output in the Americas (almost Malta 0.3 Morocco 1 084 26 twice that of the Mediterranean) provides Syria a better supply for their 762 M consumers. 206 13 Tunisia 273 8 Turkey Mediterranean citrus producers have I 482 57 clearly focused most of their efforts on Total 17 331 407 the fresh fruit market, especially oranges and lemons and recently on mandarins and other seedless easy-peelers that are • The Mediterranean area now supplies ve1y attractive to European consumers. 62% of world fresh citrus fruit expo1ts Mediterranean imports of 1.2 Mt of frozen (oranges, lemons, mandarins, etc.), as concentrated orange juice (FCOJ) and compared to 73% 20 years ago. Cuba and 0.15 Mt of grapefruits and lemons from China are newcomers to the club of non­ North and South America (chiefly from Mediterranean citrus exporters, i.e. South Brazil and Florida) compensate for the Africa, Argentina, Australia and Uruguay. low output of the Mediterranean juice • Between 1973 and 1993, Mediterranean industry. 15 Mt of fresh fruits are required Basin orange production for fresh fruit to produce this quantity of frozen con­ centrated juice. exp01ts decreased by 12%, while that of mandarins and easy-peeler hybrids increased by 112%. Mediterranean citrus • World exports of fresh grapefruits increased by 226% during the same per­ production over the last iod; USA, Cuba, Argentina and South 20 years Africa are the top producers in this area. Surprisingly, grapefruit exports from the Mediterranean citrus production for Mediterranean Basin decreased by 22%, exports and processing is summarized in despite the tristeza-free status of most Tables 3 and 4. countries in this area. General trends over the past 20 years are • Conversely, the Mediterranean Basin presented, comparing the Mediterranean has consolidated its leading world posi­ situation with that of other citrus produc­ tion with respect to exports of mandarins ing areas in the world. and hybrids (95.6% of world exports).

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 341 • e AUBERT

Table 3 World exports of fresh citrus frnits (*) (thousand t).

1992/1993 output All citrus Oranges Mandarins Lemons Grapefruits fruits & hybrids &limes

Northern hemisphere 6344 3 332 1448 765 799 USA 1106 495 20 132 459 Mediterranean region 4 960 2 666 1424 630 240 Cuba 233 130 3 100 China 45 41 4

Southern hemisphere 952 683 32 121 116 Argentina 190 73 71 46 Brazil 130 123 7 Urnguay 101 63 18 20 Australia 79 72 South Africa 452 352 30 70

World output 7296 4 015 1480 886 915

Comparison with world 1972/73 output 6 290 4 037 749 811 Mediterranean region 4 578 3 034 648 585

(*) these figures only take official fresh-fruit exporting countries into account.

Table 4 Total utilization of citrus for processing (*) (thousand t).

992/1993 output All citrus Oranges Mandarins Lemons Grapefruits fruits &hybrids &limes

Northern hemisphere 12 564 9 385 773 802 1604 USA 9 675 8 OOO 125 300 1 250 Mediterranean region 2 346 1385 338 362 261 Cuba 370 310 60 China 173 140 33

Southern hemisphere 13 014 12 354 26 456 178 Argentina 750 220 20 417 93 Brazil 11 610 11 588 22 Uruguay 45 45 Australia 366 326 6 16 18 South Africa 243 175 23 45

World output 25 578 21 739 799 1258 1 782

Comparison with world 1972/73 output 14 093 10 602 1 OOO 789 1 694 Mediterranean region 1 543 1 077 68 229 168

(*) these figures only take official fresh-fruit exporting countries into account.

342 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • ECONOMY e

Out of an overall production of 2.3 Mt of Urbanization and water resource deficits these citrus fruits in 1992/93: are seriously limiting citrus production - 1.5 Mt was exported as fresh fruit to through the loss of suitable orchard land. European markets, More than 60% of citrus trees in the - 0.34 Mt was processed for juices and Mediterranean Basin are grafted on salt­ segments, an area that was formerly dom­ tolerant Sour Orange, which unfortunately inated by Asian countries, especially is being threatened by the spread of tri­ Japan and China. In 1992/93, 44% of steza in the region. mandarins and hybrids destined for pro­ Citrus production in the Mediterranean cessing was produced in the Mediterra­ Basin will have to increase by 5.3 Mt in nean area, as compared to 6.8% in 1972. the coming 10-15 years to meet the Mediterranean and European consumers expected demand from domestic markets are generally very attracted to fresh fruit and export markets in eastern Europe. An aromas and flavours. This is a critical fac­ additional 65 M citrus trees will thus be tor for defining future strategies: needed overall. Moreover, about 30 M - for the fresh fruit market, the marketing trees should be replaced with certified season should be extended by the intro­ disease-free, tristeza-tolerant rootstock. duction of late varieties that ripen in April In all, 100 M trees will have to be pro­ and May; vided by certified nurseries. In addition to - mandarins and hybrids contain limonin the new salt- and tristeza-tolerant root­ precursors - this flavonoid can produce stocks that are needed, new cultivars will a bitter taste in heat-stabilized juices; new have to be developed to supply export technological advances are paving the markets with high quality fruits. way to novel potential uses of mandarin and hybrid juices, with new aromas and more colour; conclusion - new dual-purpose vanet1es (fresh fruit/juices) could soon be available with Biotechnological advances should help in the development of new hybrids. meeting these challenges, especially through efficient breeding and extension of elite citrus material. Mediterranean constraints Programmes to breed new cultivars for and new challenges rootstocks and scions will be introduced progressively and involve techniques that Citrus production in the Mediterranean are more specifically targeted than stan­ Basin largely relies on traditional crop­ dard breeding procedures. ping techniques. On the southern rim, New interesting traits such as tolerance to this activity is still often developed in old salinity and different viruses, insect resis­ orchards, although production systems tance and better fruit quality could soon are now being modernized in Israel, be obtained with higher efficiency. Morocco and Turkey. In contrast, mod­ ernization is quite advanced in the five Quality will also be improved by the main citrus producing countries of the development of seedless fruits (triploid) European Union (Portugal, Spain, France, with high juice content, very sweet Italy and Greece). deeply-coloured pulp and possibly lower limonin content. This new generation of Nevertheless, 60% of Mediterranean citrus cultivars will provide a basis for restruc­ orchards are owned by smallholders, with turing the Mediterranean citrus industry, an average orchard size of 0.5-2 ha; only to address the challenges that will arise 10% are citrus estates exceeding 10 ha. with the coming of the third millenium. •

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 343 • Refrigerated Fruit Juices. New Outlets for World Fruit Crops

D. LOEILLET •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR The fresh fruit juice sector, as part of the softdrin k market, 26, rue Poncelet 75017 Paris is booming: Europe is still fa r behind the USA. Th ere are many France labels but actually only two main types of juice products: freshly squeezed juices and flash-pasteurized juices. Fruits, vol. 49, n°5-6 p. 344-34 7 (English) •••••••••••••••••••••••••••••••••••••••••••••• p. 443-444 (French)

the soft drink market Germany is the European leader for all types of juices, with consumption exceed­ The European soft drink market is second ing 37 I/inhabitant/year (including former worldwide behind the United States. The East Germany). Conversely, in Portugal Japanese, despite their leading position in juice consumption is barely 3 I/inhabi­ the world economy, are still modest soft tant/year. The European average is about drink consumers, with the average con­ 18 1 (Fig. 3) sumption per inhabitant only half that of Europeans. In France, where fruit juice consumption is low (9 I/inhabitant/year), the turnover Of all beverages, the soft drink line has in this line still exceeds 5 billion francs shown the sharpest increase in sales. This and the yearly growth rates (over 25% in market has experienced 5% yearly growth volume) are promising. in the United States and Europe over the past 10 years. In contrast, tea and coffee The European fruit juice and nectar mar­ consumption is decreasing in the United ket has virtually exploded since the mid- States and that of beer is leveling off in 1970s. Sales increased by more than 60% Europe (Fig. 1) (in value) from 1987 to 1990. Western Europe consumes close to 7 Ml of fruit juices (2/3) and nectars (1/3) (Fig. 2). Imports account for 90% of these refrigerated fruit juices products consumed in Europe. are booming

In France, with annual sales growth of % annual change per person 125% between 1991 and 1992 and 80% 5 between 1992 and 1993, the refrigerated 4 fruit juice sector is expanding rapidly 3 (Fig. 4). 2 1 This small segment of the market, which 0 I has been developing in France since - 1 I I 1985, was reactivated through the intro­ duction of flash-pasteurized juices in -2 Figure 1 Soft Milk Hot Beer Wine 1990. Despite the good results over the Worldwide drink consumption -usA Europe -Japan last few years, this market only accounts patterns. 1979-1989. for 3-6% of fruit juice and nectar sales

344 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • • ECONOMY •

(depending on the data source). The Millions litres French market is still quite limited as 25 compared to the British and US markets where fruit juice consumption peak at 20 20% and 50%, respectively, with yearly growth rates varying from 10% to 25%. 15

10 a wide variety of labels 5 A public awareness campaign would be 0 - II needed to explain the broad range of fruit juice prices with which consumers are Figure 2 faced C 4 to 18 F/1). The "refrigerated" and The European softdrink market "fresh fruit juice" labels that are widely in 1989. used on packages and special offers con­ cern two entirely different types of fruit juices: fresh or freshly squeezed juices and flash-pasteurized juices (Fig. 5).

Freshly squeezed juices have a maximum West Germany . 10-day best-before date (BBD) and has to Holland . ° . be stored at 0-4 C. They are produced by Denmark . Great Britain squeezing the fruit near the sites of con­ . sumption, without adding any preserva­ East Germany Belgium tives, water or sugar. They are real fresh Spain fruit juices, which some manufacturers Ireland even prefer to call "raw" juices. France Italy I Flash-pasteurized juices are often made Greece from fruit squeezed at the production site, Portugal ' Mean EEC - followed by a heat treatment (called - Switzerland flash-pasteurization), of a few seconds at Austria ° 70 C, before refrigeration. The BBD can Sweden be extended to 24 days with this light Finland Norway heat treatment. Some brands offer juices ·-- · � of this type prepared from frozen and 0 10 20 30 40 50 even concentrated juices, which means Litres per capita/year that any fruit variety from any source can Source , FOODNEWS 1992 be used, thus reducing production costs. Figure 3 European fru it juice and nectar The full taste quality of the product is consumption in 1991. preserved in freshly squeezed juices, whereas it is noticeably diminished by the heat treatment in flash-pasteurized juices. In millions of litres 14.3 220 • In millions of francs Only freshly squeezed juices can be con­ sidered as real ready-prepared fresh fruit 8.0 130 products. These juices retain the organo­ leptic features of the fruit, which are close 62 to those of fruit juices that are squeezed 3.6 just before consumption. Figure 4 Variations in refrigerated fruit Further down the fruit juice taste quality 1991 1992 1993 juice sales in large and small scale come pasteurized juices, and then I French supermarkets fro m Source : Lineaires juices prepared from frozen concentrates. 1991 to 1993.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards .• 345 • e LOEIILLET

packaging 40 different orange juices, 10 of them were found to have suspiciously high Expansion of the fresh fruit juice market sugar contents. is dependent on the development of packagings that guarantee total product protection. Packaging specialists regularly offer different package closing systems important constraints and new materials. There are endless Fruit producers and manufacturers who innovations, e.g. a complex external are obliged to provide their clientele with packaging system for double cap protec­ fresh fruit juices of consistent taste quality tion, a perfectly watertight retractable year-round are faced with major con­ su·aw, the novel use of glass bottles for straints. It means working throughout the flash-pasteurized juices, and a wide range year with a limited number of fruit varie­ of container sizes. ties, and the need for consistency makes supply management extremely difficult. Some manufacturers overcome this con­ health and natural products straint by freezing freshly squeezed juice All surveys confirm that the consumer so that it can be stored for months. craze for these new juice types is closely Others are conducting joint investigations linked with the notions of dieta1y quality with producers to develop techniques for and natural products. lengthening the crop period or maintain­ ing the fresh fruit under cold storage con­ The French Direction Generale de la ditions. Concurrence, de la Consommation et de la Repression des Fraudes (DGCCRF) has In Australia, Valencia orange producers been surveying the market since 1992 to go as far as delaying the crop period by detect occasional frauds. Hence, in 4 months. However, this extension affects November 1992, the magazine Que the next crop yields, which can be 30% choisir? published the results of a test of lower than normal. Obviously production

Fruit Juice

Refrigerated Non-refrigerated

Freshly squeezed Flash pasteurized Pasteurized BBD < 10 days BBD > 24 days BBD several months

Ready-prepared 100% Juice from 100% Juice from product pure juice concentrate pure juice concentrate

Ulti, Fruvita, Delios, St Ivel, "Mon Verger", Tropicana, Today's, Juice Bowl, Freelin Pamp1yl, Stivel Muller

Figure 5 Flowchart for various types of fruit juice. ++++ Scale of Quality 880: best-before date.

346 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • ECONOMY •

costs are markedly affected by such pro­ these products can be found in stores duction decreases. (ready-prepared fruit and vegetables, dairy products or drinks departments). Cold storage of fruit with a suitable fungi­ cide can lengthen a fruit's storage life Producers with high production costs or from 6 to 12 weeks postharvest. low tonnage output can no longer be competitive on the same markets as large Production costs for refrigerated fruit international fruit juice producers. They Orange 66% juices are much higher than that for long­ must resolutely turn towards this very life juices, i.e. from 50% to 100% higher, quality-oriented fruit juice market. depending on whether the juice is flash­ pasteurized or freshly squeezed. The outlets handle the refrigerated fruit juice market and the frozen fruit juice market developed by some Israeli and prime opportunities French manufacturers.

Manufacturers were forecasting a 20-40% Orange and grapefruit juices account for r more than 80% of the market, followed sales increase for f uit juices in France for by fruit cocktail, grape and apple juices Source : Lineaires 1994. These growth rates are quite high, (Fig. 6). Some manufacturers have placed even though they are 4- to 5-times lower all of their hopes on refrigerated pineap­ Figure 6 than the rates obtained since 1991. Some ple juice, which has not yet been avail­ Refrigerated fruit juice. distributors predict a leveling-off period able. Breakdown of the Fre nch which will enable product consolidation. market on a flavourbasis. Innovative producers and food manufac­ The wide variety of fruit juice labels, such turers still have many options available, as "fresh", "freshly squeezed", "pure which should allow them to overcome juice", "multivitamin" and "light pasteur­ the handicaps that restrain this promising ized", has led to marked differences in sector. • selling prices and in the locations where

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 347 Publication de !'O bserw11.oire des marches du CI RAD-FLHOR

EDITORJAf.

J. GANRV SOMMA!RE lA QUAUTE : UN ENJEU ECONOMIQUE, UNE PROBLEMATIQUE Over the years, SCIENTIFlQUE Le CIRAD-FLHOR, p. 2115 RF.VUE DE 'l'oul conune la produc CIRAD-FLHOR has PRESSE tivite agricole pow-Jes derniE!res d6ccnnies, nnt:llNA'TIONALE la qualitc des produits agro-alimentaires constituera un enjeu au travers de son F. J'AJA.C majeur des anneeo a venir. La saturation des marches agricoles, steadily increased its couplee a l'Cvolution des modes de vie et des exigences des "Observatoire des Marches », p. 81:t 7 NOUVBLLES DU consommateurs, donnent unc nouvelle dimension a la information gathering MARCHK FRANCAIS problematiquc de la qualite. Mais qu'entcnd-on exaclem�nl P_ar a su acquerir et augmenter qualite, alors quc derriE!rece termc se cache wie grande d1v�1te de significations ? La qua.lite est definie comme l!tant !'aptitude potential through the p. 8814 DOSSIER DU MOIS d'un bien ou d'un service 8. satisfaire Jes besoins exprim6s ou au fiJdes ans un potentiel D. LOElLLET potcntiels des utilisateurs. Il s'agit d'unc evolution recente qui "Market Observato1y". tend plus a considerer la fonction, le client et l'organisation que informationnel en matiere GONSt:JIVt'.S O'Al'ANAS le produit scul. L'enjeu est important cur on peut estimel' que In March 1994, UN MARCHt: P.l'ISURSIS l'l!conomiel!t la politique agricolcs sont entrCcs maintcnant dans d'economie. Pour qu'un un£! veritable dynamique de qualitC, et qu'il s'ugit 18. d'Wle nouvelle voie de c.roissancc et de �ponse 3 la crise des dCbouch6s Fru iTrop, a monthly p, 16 ANANAS DE et des rcvenu.s 11gricoles. Pour Jes productions fruitieres et plus large public en soit COTE·D'l\'OIRE horticoles des TI!gions tropicales et mediterraneenncs, cette information trade Cvolution re<:ente prcnd wie dimension toute particuliCre, en beneficiaire, un bulletin raison des produits cux.mc!mes, sou vent. Cvocatcurs de nouvenuM bulletin on tropical p. 16 &: 17 ���:si�:g�TION et d'cxotisme, et de la nature des marches com;id61'6S, qu'Hs mensuel d'information LA Uf.CLEMt.'ITA1lON soicnt limitCs aux consommations locales ou destines a r COMPIJ;"!'t: Ofi \'exportation. Les mots d'ordre f>O!lt : differenciation, fruits and f esh and t.'OCM 8A!'IAl'1F. normalisation. certification Ils'agit maintenant de s'int.cresser sur Jes flux commerciaux a la " coru.truction sociale de la qualit.e " qui doit N:re processed citrus bas " negociCe " avec tous Jes acteun: de la fllihe. C'ebt dans un lei 18 19 )fERCURIALES des fruits tropicaux et des p. et JWROPEEN'NES etat d'esprit que chereheurs et part.cnaires socio-professlonncls been launched so ont abordl! cctte problcmntique de Ja qualite, a l'occnEion des agrumes frais et transformes, jourllees de septembre au CffiAD·l< 'LHOR 8 Montpe\lier, pour p. " AGENDA proposer des voics de recherehe permettnnt de rt'ipondrc nux that more people can ANNONCES ertjeu.x de demain. • FruiTrop, a ete lance. benefit from our Le premier numero a paru analyses. en mars 1994. FruiTrop provides milliards de lires : depenses des fa milies italiennes Fru iTr op propose aux the various operators en /994 pour J'achat de fruits et legumes differents interlocuteurs in the sector with de la filiere une information regular, reliable fiable et reguliere au niveau Septembre 1995 . 1 7 . page l information on trade Version fran9aise des echanges, des prix en movements, European Europe, de la prices, regulations and, in addition, an overview reglementation, avec, en outre, un tour d'horizon a of news on tropical fruits around the world. !'echelon mondial de l'actualite en matiere de fruits Each monthly issue includes a close-up on tropicaux. Chaque mois, un dossier complet presente un a particular fruit or country and a technical sheet fruit ou un pays particulier. Une fiche technique vient in each month's issue. s'ajouter aux autres donnees du bulletin. Two editions of FruiTrop are published, one FruiTrop est edite sous la forme de deux fascicules, dont in English and one in Frenc h. There are 11 issues l'un est reclige en frarn;:ais et l'autre en anglais. per year. Annuellement, paraissent 11 numeros.

Subscription Bulletin / Bulletin d'abonnement

I wish to subscribe for one year to FruiTrop Name/ Norn : Organization: / Societe : Title / Fonction : Adress: Adresse : Town / Ville : Postal Code / Code postal Country/ Pays : Phone : Fax : VAT Number :

Prepayment required / Prepaiement demande - Subscription Rate / Tari! Abonnement : 11 issues per year / 11 numeros par an : 750 F HT Payment / Reglement effectue 0 by check to the order of CIRAD-FLHOR (IRFA) / par cheque a l'ordre du CIRAD-FLHOR (IRFA) 0 by transfer to the account of CIRAD-FLHOR / par virement au compte du CIRAD-FLHOR Bank Code / Code banque : 30004 - Branch Code / Code guichet : 00892 - Account n° / N° Compte : OOO 213 600 92 Banque nationale de Paris, 51, avenue Kleber, 75116 Paris, France. II. Cropping techniques

Table Grape Growing in Tropical Areas. V. DELAITRE et al. [ Vitis vini/era, dessert grapes, grapevines, tropical zones, varieties, flowering, yields, production cycle, 350-352 agricult.ural economics, production costs, Guadeloupe] [ Vitis vinifera, raisin de table, vigne, zon tropicale, variete, floraison, rendement cycle de developpemem, economie agricole, coGt de production, Guadeloupe) [Vitis vinifera. uvas de mesa, vid, zona tropical, variedades, floraci6n, renclimiento, ciclo vital, economfa agricola, costos de producci6n, Guadalupe]

On-site Topworking of Guava Trees. J.P. LYANNAZ [Psidium guajava, topworking, production, cost analysis. Guadeloupe] 353-354 [Psidium guajava, surgreffage, production, analyse des coGts, Guadeloupe] [Psidium guajava, sobreinjertos, proclucci6n, analisis de costos, Guadalupe)

Floral Induction Study in Mango in Guadeloupe. J.P. LYANNAZ [Ma11g(fera indica, induced flowering, processing, potassium nitrate, application rates, 355-358 treatment date, Guadeloupe) [Mangijera indica, floraison induite, traitement, nitrate de potassium, close d'application, date de traitement, Guadeloupe] [Mang/fem indica, floraci6n inducida, procesamiento, nitrato de potasio, dosis de aplicaci6n. fecha de tratamiemo, Guadalupe)

Hand Pollination in Sugar Apple. X. CoGEZ et al. [Annona squamosa, pollination, fertilization, frn iting, weight gain, Guadeloupe) 359-360 [Annona squamosa, pollinisation, feconclation, fructification, gain de poids, Guadeloupe] [Annona squamosc1, polinizaci6n, fecundaci6n, fructificaci6n. ganancia de peso, Guadalupe]

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 349 Table Grape Growing in Tropical Areas

V. DELAITRE, J.P. LYANNAZ •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR Promising results have been obtained in preliminary table grape Station de Neufchateau Sainte-Marie cropping experiments in Guadeloupe. Va riety trials highlighted 97130 Capesterre-Belle-Eau the interesting qualities of cv Muscat d'Alexandrie. Guadeloupe

Fruits, vol. 49, n°5-6 •••••••••••••••••••••••••••••••••••••••••••••• p. 350-352 (English) p. 446-447 (French)

introduction Seven different table grape varieties were studied on various rootstocks and cut­ Table grapes are being grown to an tings: increasing extent in tropical regions - Cardinal (on RllO, S04, 3309c), worldwide (Thailand, Brazil, Venezuela, - Dattier de Beyrouth (on S04, cuttings), Colombia, etc.). In the West Indies, this - Muscat cl'Alexanclrie (on RllO, S04, cut- crop could be interesting for diversifica­ tings), tion and reducing grape imports (about - Italia (on Rl 10, S04, cuttings), 400 t in 1992). - Seedless (on S04, cuttings), - Centennial (cuttings), A table grape research programme was -Thompson Seedless (on S04, cuttings). set up in July 1990 at the CIRAD-FLHOR research station in Vieux-Habitants, Guadeloupe. A 1-year vine tying opera­ results tion was first undertaken, followed by an Experimental plots were cliviclecl into efficient crop protection programme that three parts to determine the best produc­ began in 1993. The prelimina1y results are tion periods: the first was pruned in ve1y encouraging. December and June; the second in February and August; and the third in · ·: ·, April and October. This test was con- material and methods ducted from December 1991 to August 1992. The bud burst rate after pruning, Two 0.3 ha experimental vineyard plots flowering rate and grape weights and were planted in July 1990 on the same quantities were determined. site with two diffe rent irrigation systems: Because of pest damage to the crops, - trickle system: 2 tricklers, 4 1/h/plant, trials focusing on pruning elates were - sprinkler system: 1 sprinkler, 35 1/h based on flowering rates rather than /plant. grape yields. Two plant training systems were tested at the same planting rate of 1600 / ha bud burst rates (2.5 m between plants within the row and To determine optimal pruning lengths, between rows): spurs (branches with two buds) were - espalier training: vines tied vertically considered separately from long branches (bilateral cordon), C 4-6 buds). For practical reasons, meas­ - arbour training: vines tied to arbours urements were only clone on espalier­ (2 m high). trained vines.

350 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROPPING TECHNIQUES •

Higher bud burst was obtained on spurs Floriferousness seemed to be best favou­ than on long branches, as noted in red on the Rl 10 rootstock. Table 1. In Guadeloupe, the best flowering could This indicates that it would be generally thus be obtained with cv Muscat best to prune branches to the two-bud d'Alexandrie, pruned to the two-bud length in order to obtain optimal bud length in February and August. Under burst. Moreover, the best bud burst see­ these conditions, the vines are leafless med to occur after a drought period in during the cyclone season and fru itless February-March. during the rainy season. flowering rates yields As shown in Table 2, the highest fl ower­ Despite the fact that a large proportion ing rate was obtained following February of the April and June grape crops were pruning. The same rates were obtained destroyed by a pyralid moth, almost on spurs and long branches. Cultivars 6000 kg/ha of cv Muscat d'Alexandrie Muscat d'Alexandrie, Cardinal and Ruby grapes were harvested. These initial Seedless bore the most flowers. results are ve1y promising (Table 3).

Table 1 Bud burst rates for seven grape varieties grown on espalier at Vieux-Habitants, Guadeloupe (n° burst buds/total n° buds).

December pruning February pruning April pruning Mean/cv Spur Long branch Spur Long branch Spur Long branch Spur Long branch Mean

Ruby/ cutting 64 39 87 89 92 85 81 71 76 Cardinal/3309 63 42 81 62 93 77 60 60 60 Centennial/cutting 38 49 85 71 89 71 71 64 67 Thompson/S04 86 32 92 78 93 84 90 65 77 Italia/110 71 55 67 71 100 73 79 66 72 Italia/S04 65 88 83 88 67 80 75 77 Dattier/cutting 80 56 87 76 96 72 88 68 78 Dattier/S04 67 93 67 84 50 81 58 69 Muscat d'Alexandrie/110 58 20 79 67 92 82 76 56 66 Muscat d'Alexandrie/S04 66 35 81 61 90 25 79 40 59

Mean 66 41 84 72 92 69 Monthly rate 63% 78% 80%

Table 2 Flowering rates for seven grape varieties grown on espalier at Vieux-Habitants, Guadeloupe (n° burst buds/total n° buds).

December pruning February pruning April pruning Mean/cv Spur Long branch Spur Long branch Spur Long branch Spur Long branch Mean

Ruby/cutting 27 16 32 27 10 11 23 18 20 Cardinal/3309 18 25 45 25 30 33 31 28 30 Centennial/ cutting 12 0 4 8 4 8 6 5 5 Thompson/S04 6 0 19 18 6 11 10 3 6 Italia/110 10 0 13 0 17 13 13 4 8 I.talia/S04 6 14 0 0 0 7 0 3 Dattier/cutting 11 6 15 22 3 5 10 11 10 Dattier/S04 10 4 0 6 0 7 0 3 Muscat d'Alexandrie/110 17 0 46 45 43 30 35 25 30 Muscat d'Alexandrie/S04 14 50 44 55 45 25 39 43 41 Mean 14 12 24 20 16 14 Monthly rate 13% 22% 15%

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 351 • • DELAITRE and LYANNAZ

Table 3 Grape yields in June 1992 for espalier-trained vines at Vieux-Habitants, Guadeloupe.

No grape Mean bunch Yield bunches/ plant weight (g) (kg/ha) Ruby/cutting 8 219 3 540 Cardinal/3309 Centennial/cutting Thompson/S04 222 710 Italia/110 2 200 666 Italia S04 2 187 533 Dattier/cutting 4 244 1 567 Dattier/S04 3 219 1 243 Muscat d"Alexandrie/110 13 266 5 762 Muscat d'Alexandrie/S04 9 247 3 857 Mean 5 225 2 235

length of the production cycle Grape prices are generally below 12 F/kg in Guadeloupe (minimum import price). The mean interval between pruning and Considering the technical constraints that harvest was 140 days. The earliest cultivar necessitate production in June and was Muscat d'Alexanclrie (125 clays), and December, note that: the latest were Italia and Dattier de - in June, grape consumption is quite low Beyrouth (150 days). (20 t in 1991), whereas prices are high (20-25 F/kg in 1991); - in December, grape consumption is economic data high (100 t in 1991), whereas prices are the lowest (12 F/kg). table grape imports in Guadeloupe In Guadeloupe, the grape production Grape imports increased by more than breakeven point (10 F/kg) is reached 100 t in 1991, reaching a total of 400 t in when a minimum of 7600 kg/ha/year are 1992. This could be explained by a mar­ produced on espalier-trained vines. ked rise in grape imports from Chili and Arbour production data are still insuffi­ Spain, whereas imports from metropolitan cient to reach a final conclusion. France dropped drastically (- 100 t in 1991 ). These imports were ve1y high in the conclusion September-December and March-April An efficient crop protection programme periods. Wholesale grape prices in was set up in 1993 and the interesting Guadeloupe were very high in July qualities of cv Muscat d'Alexandrie were (35 F/kg in 1991) and levelled off at highlighted. 10-15 F/kg throughout the rest of the There are still substantial technical year. choices to be made, especially to deter­ mine the best rootstocks, vine training production costs and pruning techniques under local con­ Annual vineyard installation and mainte­ ditions. nance costs were calculated after a year Although tropical grape production is of grape production at Vieux-Habitants labour intensive and requires high initial (Guadeloupe): investments, it could be cost effective in - installation: 187 OOO F/ha for espalier Guadeloupe. Thirty ha of vineyards could training and 237 OOO F/ha for arbour train­ produce enough grapes to meet the ing, domestic market demand. • - maintenance: 82 OOO F/ha/year for espa­ lier and 128 OOO F/ ha/year for arbour.

352 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards On-Site Topworking of Guava Trees

J.P. LYANNAZ •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR A technique to double-graftguav a trees (cv Beaumont) could provide Station de Neufchateau Sainte-Marie an alternative to standard uprooting/replanting practices. 97130 Capesterre-Belle-Eau Guadeloupe ••••••••••••••••••••••••••••••••••••••••••••••

Fruits, vol. 49, n°5-6 p. 353-354 (English) p. 448-449 (French)

introduction gated. Local and poor-performing impro­ ved varieties could be replaced by an on­ The development of small- and medium­ site topworking technique, as an alterna­ sized fruit processing units and the boom­ tive to conventional uprooting/replanting. ing tourist industiy have promoted an increase in the demand for local fruit in Guadeloupe. Local guavas are in high material and methods demand since imported guava pulp is often expensive and of unsatisfactory topworking technique quality. Topworking is carried out as follows: The guava cultivar Beaumont is of interest for its high yields and quality, particularly 1. The trees to be grafted are cut back to with respect to its processing perfor­ the main branches, and the remaining mance (colour, flavour, sweetness and trunk and main branch bases are daubed with thick whitewash to hinder sun Figure 1 acidity). Varietal conversion of a guava Patch budding. orchard with this cultivar was thus investi- scorching. 2. Four to six vigorous ratoons are selec­ ted; they should be spatially well distrib­ uted to give a well-balanced architecture : , . t I�. /jl to the tree. Frail, poorly located and ' 'I ,· '. -i·, , ! t' , n., . , • ·,I I . excess branches are pruned. l. lr , ·,'., I;, : l' � ( ·"./. 3. Cultivar Beaumont is grafted onto the stocktree by patch budding when the ratoons are 10-15 mm diameter and beginning to lignify (Fig. 1) The buds are protected with grafting plastic. 4. Fifteen to twenty clays after grafting, ties are removed from the bud union on grafted branches and they are partially Scion cutting Prepared Rootstock/scion pruned to 10-15 cm above the graft. Bud rootstock assembly sprouting occurs veiy soon after this operation C < 1 week).

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 353 • • e LYANNAZ

5. Branches are pruned to about 2 cm These results indicated that topworked above the scions after they have reached guava trees began producing slightly soo­ a length of 15-20 cm, and the pruning ner than ungrafted trees. wounds are sealed with grafting wax. Finally, a cost-effectiveness analysis study 6. Buds are removed from the rootstock showed that the cost of labour involved on a regular basis, generally 2-3 times. in the topworking procedure (i.e. prun­ ing, disposing of wood left after pruning, experiments whitewashing trunks, grafting, bud remo­ val, partial and final pruning of grafted The double-grafting experiments were branches) was equal to that required for carried out at the Vieux-Habitants conversion (uprooting, purchasing new research station, Guadeloupe, in an irri­ plants, replanting) of a similar-sized plot, gated 0.3 ha guava orchard cropped with i.e. about 11 OOO FF/ha. various cultivars that have poorer yield performance than cv Beaumont. The trees to be grafted (3.5 years old) were cut back on 27 October 1988, and grafted on conclusion 14-15 February 1989. They were planted The present study highlighted the benefits in stands of 357 trees/ha (7 m x 4 m). of using the topworking technique to An equivalent-sized control plot (0.3 ha) rehabilitate a guava orchard as compared was planted with cv Beaumont cuttings in to an uprooting/replanting procedure, i.e. 1989. the onset of production is earlier for the same technical costs. The 1992 harvest was poor because of a fungal disease and results and discussion the yield figures will therefore have to be confirmed. An assessment carried out in late March 1989 revealed a success rate of more than With this topworking technique, an effi­ 95% in the double-grafted orchard. In cient experienced grafter is required to December 1989, the first blossoms were obtain a high grafting success rate in a noted, with subsequent yields of 11.6 t/ha single operation. in 1990, 32.6 t/ha in 1991, and 32.5 t/ha Another interesting advantage of this in 1992. However, 50% of the guavas technique is that it avoids potential rot could not be harvested because of a fun­ development due to inefficient uprooting; gal disease (Phoma sp.). this is a common problem after uprooting The yield results on the control plot were and land clearing operations. • 9.1 t/ha in 1990, 24.8 t/ha in 1991 and 59.5 t/ha in 1992.

354 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards Floral Induction Study in Mango in Guadeloupe

J.P. LYANNAZ •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR Mango trees were treated with several flower-inducing substances Station de Neufchateau Sainte-Marie by leaf spraying and undertree irrigation applications. The initial 97130 Capesterre-Belle-Eau results highlight the promising mango cropping potential in Guadeloupe the Lesser Antilles. Fruits, vol. 49, n°5-6 •••••••••••••••••••••••••••••••••••••••••••••• p. 355-358 (English) p. 450-452 (French)

introduction out from 1970 to 1978, and yields were thus increased threefold (ANONYMOUS, Two mango cultivars ( and ) 1978). from Florida (USA) were introduced in Several other countries, particularly Guadeloupe in the early 1980s with the Mexico, have obtained positive results by aim of developing the export mango adapting these techniques to their local industry. However, some flowering prob­ conditions (NUNEZ-ELISEA, 1986 a & b; lems were noted when these trees were NUNEZ-ELISEA & CALDEIRA, 1987). grown under prevailing conditions in Guadeloupe. Flowering was found to be In the present study, two different flower poor and highly irregular (December to induction treatments were conducted with April flowering period) in lowland areas. these substances in a CIRAD-FLHOR These cultivars also show quite marked research orchard at Vieux-Habitants biennial bearing. (Guadeloupe). Irrespective of the varietal factor, this situ­ ation could be explained by several cli­ material and methods matic conditions: - irregular and often insufficient d1y per­ comparison of two flower-inducing iod, substances on cv Eldon mango - high humidity, - high mean temperatures year-round. Two flower-inducing substances, Flowerset (Bayer) and Miracle Blum The fumigation technique has long been Powder (Philippine Orchard Corp.), were used in India and the Philippines for flo­ tested in Janua1y 1991 on cv Eldon man­ wer induction. By this technique, trees gos grafted on 10 year old non-improved are fumigated for several days until the local mango trees. appearance of ; the opera­ tion is repeated 2 months later if no flo­ Flowerset (240 g potassium nitrate/I) was wer bud burst occurs within 15 days. The applied by leaf spraying at 15 cc/I (36 g/1 results, however, are often poor and the postassium nitrate). operation is expensive and labour inten­ The chemical composition of Miracle sive. Blum Powder is not given, but analyses Chemical treatments were successfully indicated a total nitrogen content of 25%, tested at the beginning of the 1970s. In 14.5% as ammonium and 10% as nitrate. the Philippines, for instance, potassium Part corresponds to ammonium nitrate, nitrate spraying treatments were carried and the rest (about 1/3 of the ammonia

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 355 •• e LYANNAZ

form) to other salts (phosphates, sul­ mango and other fruit trees, was used as phates, chlorines, etc.). It was applied at a control compound for the potassium 15 g/L nitrate treatments (V ILLAUME, 1987; V ILLAUME & NYEMBI, 1991; YOON et al., These substances are often used instead 1989). of raw potassium nitrate in the Philippines since, for milita1y security rea­ Four different treatments were carried out sons, it is sometimes difficult to obtain with three flower-inducing substances: authorization to use the latter compound - Flowerset (240 g/1 potassium nitrate): (ANONYMOUS, 1978) 15 cc/I by leaf spraying (10 I/tree), - potassium nitrate: 10 and 40 g/1 by leaf For each treatment, even trees were trea­ spraying (10 I/tree), ted by leaf spraying (about 14 I solu­ - paclobutrazol: 60 cc Cultar (IC!) in 20 1 tion/tree). For practical reasons, only the water by undertree irrigation followed by lower 2/3 of the foliage was treated. sprinkling. This test was carried out along a uniform comparison of three flower-inducing line, at two trees/plot, two replications substances on cv Haden mango and adjacent controls. Since the prelimina1y flower induction tests with cv Eldon were successful, a sec­ results ond experiment was carried out in March 1991. However, only nitrate treatments comparison of two flower-inducing were undertaken because data for the Miracle Blum Powder trial was not yet substances on cv Eldon mango available. Figure 1 shows the positive effects of two 3 year old cv Haden mangos graf­ flower induction treatments as compared ted on local non-improved mango trees to the control: after the natural flowering \N ere treated in this second experiment. stage (T + 15 cl ) in controls and treated Terminal buds on all trees tested were at trees, flowering began on the treated Figure 1 the swelled or dormant stage. trees (peak at T + 35 cl); the best results Flower induction treatment were obtained with Miracle Blum on cv Eldon mango: mean Paclobutrazol, a growth regulator com­ number/tree. monly applied to increase yields in Powder. The flower-inducing substances markedly increased yields, as measured by the har­ 100 vested fruit quantities and weights (Figures 2 & 3). 90 However, there were no differences. bet­ 80 ween the three treatments when separate

356 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROPPING TECHNIQUES 8

discussion 12 effects of nitrates 10 In both experiments, potassium nitrate was found to induce flowering when applied by leaf spraying at a specific 8 physiological stage: swelled and dormant tem1inal buds and brittle dark-green ·v 6 ,,._ leaves, corresponding to a branch age of 6-7 months. 4 V Ammonium nitrate was also found to induce flowering in the Miracle Blum Powder treatment. However, the exact 2 role of this compound would be difficult to determine because of the complex 0 chemical composition of this product. 24 19 20 21 22 23 25 26 27 The present results generally confirmed Weeks post-treatment the hypothesis of NUNEZ-ELISEA (1986) that nitrate ions have a prime role in the ---•-- Control --II- Flowerset • Miracle Blum chemical flower induction process in mango. effect of paclobutrazol Figure 2 Fl ower induction treatment potassium nitrate treatment dose The ve1y poor results obtained with the on cv Eldon mango: mean fruit paclobutrazol treatment could be explai­ weightjtree. The results indicated an ideal potassium ned by several factors: nitrate treatment close of 10 g/1, applied - the young age of the plants treated in March, to obtain flowerinduction in cv (3 years), Haden under the prevailing climatic con­ - the low paclobutrazol sensitivity of the Figure 3 ditions of Guadeloupe. varieties tested, Fl ower induction treatment - the low late-season flower-inducing effi­ on cv Eldon mango: mean fruit In the Colima region of Mexico, a dose of ciency of paclobutrazol. number/tree. 40 g/1 has been recommended for cvs Haden and Manila with treatments begin­ ning in the first 2 weeks of November. Jn 25 the Philippines, a dose of 10 g/l is effec­ tive for cvs and with treat­ ments in February or October. 20 The treatment dose can therefore be adjusted according to varieties, and the application date should be set in terms of the physiological stage of the branches. application dates

Treatments in mid-Janua1y on cv Eldon 5 and mid-March on cv Haden led to an increase in natural seasonal flowering and induction of late season flowering. 0 These techniques should now be assessed 24 19 20 21 22 23 25 26 27 in terms of reproductivity and the effects Weeks post-treatment of treatments relative to the physiological stages of branches and climatic condi­ ---•- - Control - Flowerset ---•-- Miracle Blum tions.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 357 • e LYANNAZ

conclusion Table 1 Results of the flower induction treatment with cv Haden. Th se initial positive results could lead to more harmonious and reliable develop­ Flowers/ N° fruits/ Fruit weight/ Size ment of mango cropping in Guadeloupe tree tree tree (kg) and throughout the lesser Antilles.

Control 1.76 0.94 0.37 0.39 Although further experiments should be KN0 10 g/l 59.25 25 7.16 0.29 carried out to improve cropping tech­ 3 niques, under the treatment conditions KN0 40 g/l 32.75 14 4.3 0.31 3 investigated in this study, it seems pos­ Cultar 0 0 0 0 sible to obtain: Flowerset 77 32 10.76 0.34 - more uniform flowering, - a reduction in biennial bearing, - late season flower induction. In these conditions, it would be interest­ However, in th light of the present ing to compare the growth-reducing know-how, these techniques will prob­ effects of paclobutrazol with the flow r­ ably not enable year-round mango pro­ inducing effects of nitrates (VooN et al., duction under the highly irregular tropical 1989) climatic conditions in Guadeloupe. •

references

ANONYMOUS, 1978. VUILLAUME C., 1987. The Philippines Recommends for Mango, Paclobutrazol - PP 333. Un nouveau regula­ 70 p. Los Banos, Laguna, Philippines, teur de croissance et des resultats promet­ Philippine Council for Agriculture and teurs en cultures fruitieres tropicales. Ressources Research, p. 11-13. Montpellier, France, RA 1987, IRFA/CIRAD, 4 p. {document n· 100). NUNEZ-ELISEA R., 1986a. Producci6n temprano de mango 'Haden' y VUILLAUME C., NYEMBI Z., 1991. 'Manila' con aspersiones de nitrato de pota­ Vers une maTtrise de la floraison du man­ sio. Campo Agricola Experimental Tecoman, guier au Cameroun. Utilisation d'un regula­ Mexico, SAEH-INIFAP-CIAPAC, 8 p. teur de croissance : le paclobutrazol. Fruits 46 (2) : 187-198. NUNEZ-ELISEA R., 1986b.

Potencial del nitrato de amonis (NH4 N03) VOON C.H., PITARPAIVAN C., TAN S.J., 1989. para inducir la floraci6n del mango. In: Mango cropping manipulation with Cultar. In: Congresso nacional de la ciencia del suelo, 3rd International Mango Symposium. Acta 19. Resumen. Sociedad Mexicana del suelo, Horticulturae 291, p. 219-228. p. 29.

NUNEZ-ELISEA R., CALDEIRA M.L., 1987. Adelanto de la floraci6n y cosecha en mango 'Haden' con aspersiones de nitrato de amo­ nio. In: IX Congresso brasileiro de fruticul­ tura. Sociedade brazileira de fruticultura, p. 561-566. ••••

358 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards Hand Pollination in Sugar Apple

X. COGEZ ••••••• •••••••••••••••••• • • ••••••••••• •••••• •• Palmiste 97113 Gourbeyre A simple hand pollination technique is recommended to enhance Guadeloupe sugar apple productivity, a fr uit that is highly appreciated J.P. LYANNAZ in Guadeloupe. CIRAD-FLHOR Station de Neufchateau •••••••••••••••••••••••••••••••••••••••••••••• Sainte-Marie 97130 Capesterre-Belle-Eau Guadeloupe

Fruits, vol. 49, n°5-6 p. 359-360 (English) p. 453-454 (French)

introduction The method was simple: pollen was col­ lected from flowers at the male stage Natural pollination rates are low in many (fully separated ). It was then dus­ Annona species, thus explaining their ted onto the stigma of other flowers in poor productivity. the female stage (barely half-open petals) with a small fine-bristle brush. Trees were flowers are highly protogynous Annona monitored for fruit setting 10-15 days (i.e. the stigma are receptive before sta­ postpollination. men maturity), which limits self-pollina­ tion. The morphology of the flowers This operation was repeated four times at (colour, shape) also reduces entomophi­ 15 day intervals in order to treat as many lous and anemophilous cross-pollination. flowers as possible. Many cherimoya and/or atemoya produc­ Flowers on cv Thai Lup were only polli­ ing regions (Australia, California, Chili, nated with pollen from the same variety Spain, etc.) have overcome this problem of flowers (intravarietal pollination). through hand pollination practices. With the New Caledonian cultivar, intra­ These techniques were modified for sugar varietal (with its own pollen) and interva­ apple (Annona squamosa), a species that rietal (with cv Thai Lup pollen) pollina­ is closely related to cherimoya and ate­ tions were carried out. moya but better adapted to dry tropical climates as found in the West Indies. results and discussion material and methods fruit settingrate Annona squamosa trees from two differ­ ent origins were used in this study: The low fruit setting rates noted in - a 4 year old tree of Florida origin A. squamosa controls were in line with (cv Thai Lup grown from seed); reported rates for A. cherimoya and - eight 2.5 year old trees of New A. attemoya, which have fruit setting rates Caledonian origin (from local improved of about 1 % under natural pollination seed). conditions.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 359 • • COGEZ and LYANNAZ

Nevertheless, sweet apple is very well Table 1 adapted to the pedoclimatic conditions of Results of hand pollination in Annona squamosa. Cote Sous le Vent (Basse-Terre) and Grande Terre. The hand pollination tech­ Treatment N. N. % fruit Mean wt/ gain vs nique boosts production and could d1us flowers set fruits budded fruit(g) control help promote sweet apple development in Guadeloupe. cv Tha'i Lup This technique, based on a mean flower­ Control 40 0 0 143 C-) ing/harvest interval of 110 days, would Intra poll. 38 38 100 230 + 62% also be interesting for grouping and pro­ Inter poll. gramming harvests. cv New-Caledonia The present prelimina1y results could be confirmed by: 4 Control 111 36 230 (**) - conducting the same study during the Intra poll. 33 30 90.9 269 + 17% whole flowering period, with a larger Inter poll. 54 50 92.6 313 + 36% sample if possible; - checking the advantages of intervarietal mean for 83 fruits harvested outside of the test period (natural pollination). pollination; .. mean for 39 fruits harvested fromcontrols and outside of the test period. - adapting the technique for other Annona species such as A. reticulata (bullock's heart), and especially A. muric­ ata (soursop); the initial results for this A fruit setting rate of more than 90% was species were quite inconclusive. • obtained in all cases with hand pollina­ tion, thus confirming the efficiency and benefits of artificial pollination (Table 1).

mean fruit weight

Very marked fruit weight gains (17-62%) further reading were obtained in all cases with hand pol­ lination. These gains were great since the AHMED M.S., 1939. mean fruit weights under natural condi­ Pollination and selection in Annona squa­ mosa and A. cherimola. Bui. Min. Agr. Egypt tions were very low. Tech. Sci. Serv. Hort. Sect, 157, 1. Mean weights of fruit produced on the GEORGE A.P., NISSEN R.J., CAMPBELL J.A., New Caledonian variety after intervarietal 1992. pollinations were higher than those pro­ Pollination and selection in Annona species duced after intravarietal pollinations (cherimoya, atemoya and sugar apple). In: (gains of 36% and 17%, respectively). International symposium on tropical fruit: frontiers in tropical fruit research, Pattaya This weight gain was associated with City, Thailand, 20-24 mai 1991. Wageningen, improved well-formed fruit (well roun­ Netherlands, ISHS, p. 178-185. ded, no flattening), which could be explained by the fa ct that hand pollina­ SAAVEDRA E., 1977. Influence of pollen grain stage at the time of tion improves pollen coverage on the hand pollination as a factor on fruit set of stigma. cherimoyer. HortScience, 12, 117.

SCHROEDER C.A., 1943. conclusion and prospects Hand pollination studies on the cherimoya. Proceedings of the American Society for Sweet apple trees are only grown on a Horticultural Science, 43, 39. single-tree basis in some creole gardens THAKUR D.R., SINGH R.N., 1965. in Guadeloupe. This fruit is therefore Studies on pollen morphology, pollination quite rare despite its popularity in and fruit set in some Annonas. Indian Journal Guadeloupe. of Horticulture, 22, 10.

360 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards 111. Genetics

GENETIC RESOURCES

Characterization and Performance of 51 Citrus Va rieties F. MADEMBA-SY et al. in New Caledonia. [citrus fruits, variety trials, environments, trees, quality, production, New Caledonia] 362-370 [agrume, essai de varietes, environnement, arbre, qualite, production, Nouvelle-Caledonie] [frutas citricas, ensayos de variedades, medio ambiente, arboles, calidad, producci6n, Nueva Caledonia]

Genetic Resources of Mangos in Cote d'Ivoire. T. GOGUEY [Mangi/e ra, germplasm, collections, hybridization, plant breeding, Cote d'Ivoire] 371-375 [Mangife ra, materiel genetique, collection, hybridation, amelioration des plances, Cote-d'Ivoire] [Mangifem, germoplasma, colecciones, hibridaci6n, fitomejoramiento, Cote d'Ivoire]

Management of Litchi Genetic Resources in Reunion. F. NORMAND et al. [litchi, germplasm, collections, provenance, selection criteria, quality, harvesting elate, Reunion] 376-382 [litchi, materiel genetique, collection, provenance, critere de selection, qualite, elate de recolte, Reunion.] [litchi, germoplasma, colecciones, proceclencia, criterios de selecci6n, caliclacl, fecha de recolecci6n, Reunion]

Inventory of Tropical Fruit Trees in Central America and the West Indies. G. BARBEAU [Anacardiaceae, Annonaceae, Guttiferaceae, Lauraceae, Lecythidaceae, Malpighiaceae, Moraceae, 383-389 Myrtaceae, Oxalidaceae, Rhamnaceae, , Rutaceae, Sapindaceae, Sapotaceae ...J [ ... fruit trees, Central America, Caribbean] [. .. arbre fruitier, Arnerique centrale, Cara'ibes] [. .. arboles frutales, America central, Caribe]

BIOTECHNOLOGY

Nuclear Genome Size Variations in Citrus. P. OLUTRAULT et al. [Citrus, , evolution, biodiversity, genomes, chromosome number] 390-393 [Citrus, taxonomie, evolution, biocliversite, genome, nombre chromosomique] [Citrus, taxonomia, evoluci6n, biocliversiclacl, genomas, numero de cromosomas]

Optimized Management of Citrus Embryogenic Cal Ii for Breeding Programmes. P. OLuTRAULT et al.

[Citrus, callus, freezing, emb1yonic development, somatic embryos, plant breeding] 394-397 [Citrus, cal, congelation, developpement emb1yonnaire, embryon somatique, amelioration des plantes] [Citrus, callo, congelaci6n, desarrollo embrionario, embri6n somatico, fitomejoramiento]

Facultative Apomixis, Spontaneous Polyploidization and Inbreeding P. OLurnAuLr et al. in Citrus volkameriana Seedlings. [Citnis, apomixis, polyemb1yony, rootstocks, isoenzymes, chromosome number] 398-400 [Citrus, apomixie, polyemb1yonie, porte-greffe, isoenzyme, nombre chromosomique] [Citnis, apomixis, poliembrionia, portainje1tos, isoenzimas, numero de cromosomas]

Protoplast Fusion in Citrus. P. OLurnAuLT et al. [Citnis, Fo 11unella, protoplast fusion, callus, somatic hybrids, isoenzymes, DNA cleavage] 401-403 [Citrus, Fortunella, fusion de protoplastes, cal, hybride somatique, isoenzyme, fragmentation de l'ADNJ. [Citrus, Fortunella, fusion de! protoplasto, callo, h1brido somatico, isoenzimas, fragmentaci6n de la AON]

Genetic Mapping of an lntergeneric Citrus Hybrid Using Molecular Markers. F. LuRo et al. [Citnts, genomes, isoenzymes, restriction enzymes, genetic maps, segregation, plant breeding] 404-408 [Citrus, genome, isoenzyme, enzyme de restriction, carte genetique, segregation, amelioration des plantes] [Citrus, genomas, isoenzimas, enzima de restricci6n, mapas geneticos, segregaci6n, fit.omejoramiento]

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 361 Characterization and Performance of 51 Citrus Varieties in New Caledonia

F. MADEMBA-SY •••••••••••••••••••••••••••••••••••••••••••••• 5. LEBEGIN Although citrus fr uits are not yet widely grown in New Caledonia, A. HAURY J.P. LYANNAZ interesting results were obtained in analyses of recently-introduced CIRAD-FLHOR grapefruit, orange and mandarin varieties. Station de Pocquereux BP 32, 98880 La Foa •••••••••••••••••••••••••••••••••••••••••••••• New Caledonia

Fruits, vol. 49, n° 5 p. 362-370 (English) introduction description of the environment p. 456-460 (French) Intensive Citrus cropping is a recent phe­ geographical location nomenon in New Caledonia (GuJLLAUMJN, 1952), and the first Citrus development New Caledonia is 18 OOO km from metro­ programme in this country only began in politan France. This archipelago is located the early 1980s (PRALORAN, 1971). Varietal north of the Tropic of Capricorn, from 19' breeding plots were set up at the CIRAD to 23· latitude S and 158' to 172' longi­ fruit research station at Pocquereux, New tude E (Map). The overall surface area is 2 Caledonia, to obtain data on the perfor­ 19 100 km , about twice the size of mance of citrus in this region, that is: Corsica and 18-fold that of Martinique. - in 1986, with 60 cultivars There is one main northwesterly/south­ - in 1991, with 160 cultivars. easterly oriented island C Grand Terre: 16 900 km2) and several smaller f1anking Interesting prelimina1y results were obtai­ islands (Iles des Pins, i:les Loyaute and 1les ned for several cultivars in an initial trial Belep: 2 200 km2) (ORSTOM, 1989). that had been planted in 1986. Grand Terre (about 400 km long, 50 km wide) has a central mountain range that peaks at 1 628 m, with deep perpendicu­ 165° Est lar valleys on the east and west coasts. New Caledonia These valleys are at low elevations C < 200 m), and longer on the west coast than on the east coast. The f1anking coral 20° Sud islands are volcanic. climate The archipelago has a tropical-to-Medi­ terranean type of climate; part of the year the territory is influenced by the inte11rop­ ical convergence zone (ITCZ), and the �Mare other part by temperate depressions from the South Pole: ° 22 Sud - mid-December to mid-April is the main hot rainy season that peaks in February­ March, tropical depressions and cyclones occur at this season; -50 km lie des Pins - mid-April to mid-May is a minor dry Tropique du Cepricorne season when the rainfall and temperature

362 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES e

decrease with the northward movement T °C mm of the ITCZ; 35 - mid-May to mid-September is the cool 500 30 450 season, which can be rainy due to the 400 arrival of cold fronts from the South Pole, 25 � 350 causing temperate depressions; 20 � 300 250 - mid-September to mid-December is the 15 main d1y season, when temperatures rise �/ 200 150 and the ITCZ begins moving southwards. 10 5 100 50 temperatures 0 � 0 1 2 3 4 5 6 7 8 91011 12 Temperature is a determining factor for Months Figure 1 citrus performance; it affects internal fruit T0 mean max. T0 mean min. Climatic data fo r La Fo a. quality and colour. 0 Elev. 18m, Lat. 21 °40'S, T mean Precipitations ° Temperatures recorded at the meteor­ Long. 169 494£. ological station in La Foa (New Caledonia) are presented in Figure 1. For T °C a 34 year period, the mean annual tem­ perature was 22.5'C (mean maximum 28 26 28.5'C, mean minimum 16.4"C) . Note that 24 these are not typical temperatures, i.e. the 22 minimums are quite low despite the aver­ 20 age to low elevation and latitude and the 18 insularity effect. 16 As shown in Figure 2, the temperatures 14 were midway between those found in 12 10 Martinique (tropical climate) and Corsica (Mediterranean climate). 1 2 3 4 5 6 7 8 9 10 11 12 Months hygrometry, evapotranspiration - San Giulano (Corsica) Figure 2 and precipitation - La Foa (New Caledonia) Comparison fo mean The mean hygrometry ranged from 48% - Riviere Lezarde (Ma1tinique) temperatures for La Fo a, Martinique and Corsica. to 96%. The potential evapotranspiration was 1463 mm, with a June low of 65 mm These soils are not a priori ideal for and a December high of 177 mm. citrus cropping because the orchards are The mean annual rainfall was 1 155 mm, flooded during the first quarter of the but there was a broad range of levels year as a result of tropical depressions. (610 mm in 1967, 2 292 mm in 1973). However, they are generally close to water sources (rivers) and more fertile Precipitation was poorly distributed than other soils in New Caledonia. throughout the year (mean 84 days), Specific development projects (banking, which means that crops have to be irri­ hilling) also enable the trees to withstand gated. temporary exundation. soil The citrus cultivars studied were planted on recent alluvial river terraces. The soils material and methods were heavy sandy clay loam (85% clay/loam). These recent alluvial soils cropping techniques have a substantial organic matter content The experimental plot, located on the (2.9%), with high magnesium content and Pocquereux River floodplain and thus only trace levels of potassium and cal­ subject to periodic flooding, was initially

cium, and no sodium. The soil pH is 5.5 a dry forest of cajeput (Melaleuca quin­ (GOOEFROY, 1990). queneruia) and candlenut (Aleurites

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •..• 363 • • MADEMBA-SY et al.

moluccana). It was cleared with a bull­ in Corsica. The trees were periodically dozer and then the soil was heavily ferti­ indexed for tristeza virus. No trees were lized (2 t/ha acetylene manufacturing resi­ positive even though tristeza has been dues1). NPK fertilizer (0-32-16; 2 t/ha) detected elsewhere in the archipelago in was applied as a basal dressing. Navel oranges from Australia that were introduced in the early 1960s. All diseased The trees were planted in turf (3 m long x trees were destroyed and an eradication 3 m wide x 60 cm high) in September campaign carried out. Toxoptera citricidus 1986. In 1992, a system of open trenches Kirkaldy, the main tristeza vector, is not was set up for quick floodwater drainage, present in New Caledonia. since such events do not last more than 12 h. With this system, the orchard was varietal performance able to withstand six 2 rn high floods Varieties were evaluated according to that occurred since the trees were IBPGR-FAO criteria and standards, the planted, without significant damage (no details are given in Appendices. The main mortality). The conditions on this varietal traits of interest for growers are: breeding plot are typical of those - fruit shape (ED/PD ratio = equatorial found in most citrus orchards of New diameter/polar diameter), Caledonia. - peel thickness (epicarp and mesocarp), The trees \N ere planted at 7 m x 7 rn - number of seeds, spacing. Each variety was represented by - juice content (%), blocks of 4 trees. Aciduous varieties - acidity (A), (lime, lemon, grapefruit) were grafted on - soluble dry extract (SDE), Citrus volkamericma, and sweet varieties - fruit weight (orange, mandarin) on Carrizo and Troyer - yield, citrange. - flowering-harvest period.

irrigation sampling Trees were irrigated with two microjets The quantitative and qualitative measure­ (36 1/h) set at 180° and on opposite sides ments for each citrus variety were based of the tree to avoid wetting the trunk. The on 10-fruit samples harvested in 1990, irrigation period lasted 5-9 months, 1991 and 1992 depending on rainfall levels, with a PE of O 75

ferti I ization results The orchard was fertilized three times grapefruits yearly, as follows: Six varieties of grapefruit (Citms paradisis 50% 1 month before flowering Quly), Macf.) were compared: Marsh SRA 120, 25% 2 months after flowering Shambar SRA 22, Redblush SRA 56, Star (September), Ruby SRA 199, Ruby SRA 286 and 25% 4 months after flowering Thompson SRA 121 (Photo 1). (November) . canopy The annual fe rtilizer input for 7-year-old trees was 1500 g urea and 2500 g NPK Grapefruit trees in New Caledonia have a (13-13-21). smaller canopy volume than trees grown under real tropical climate conditions plant material (only 42 m3 for 7-year-old trees, 4.1 m high x 4.6 m diameter canopy). All six (1) Type of whitewash health status varieties had a rounded habit. containing 13% Cao, obtained About 60 citrus varieties were grown on following water erosion of the test plot; the results for 49 of these ripening period calcium carbide. Citrus growers can obtain this product free-of­ are presented here. All plant material was The grapefruit varieties were late, with a r charge from the acetylene certified and originated f om the INRA­ 9 month period from flowering manufacturer. CTRAD agricultural research station (SRA) (September) to maturity Qune). The fruit

364 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES •

can stay on the trees for 4 months with­ out any change in internal fruit quality. fruit quality The fruit was round (ED/PD = 1/1), with a thick skin (9 mm) and few seeds (n = 5). The soluble dry extract levels were low with little variation (SDE = 7.5). The varieties differed in terms of acidity and juice content. Shambar and Ruby were the least juicy (45% and 48%, respectively) and the most aciduous (A = 1.1 g and 1.07 g citric acid, respec­ tively) of all the grapefruit varieties. Star Ruby was the most juicy (57%) and least aciduous (A = 1.03 g) variety. Grapefruits grown in the New Caledonian climate had ve1y little fruit colour. The temperatures were too low during ripen­ ing to produce the pigments responsible for colouring in grapefruit (lycopene and Photo 1 beta-carotene). Hence, cvs Redblush, The six varieties of grapefruit Shambar and Thompson produced fr uit Table 1 studied in New Caledonia. with lower colouring than fruit grown in Cumulated yields/tree for grapefruit tropical conditions. Star Ruby was the varieties over 4 years. only variety that produced suitably colou­ red fruits under the conditions investi­ 4-year cumulated gated in the present study. Variety yield (kg/tree) Shambar SRA 22 727 production Marsh SRA 120 750 Thompson SRA 121 691 There was little variation in grapefruit Ruby SRA286 678 weights (mean 510 g). The largest fruits Star Ruby 199 567 were produced by cvs Ruby and Redblush SRA 56 488 Redblush. Mean fruit yields/tree for all varieties were 105 kg at 5 years, 130 kg at 6 years, 167 kg at 7 years and 249 kg at oranges 8 year, for a mean cumulated yield/tree Fourteen varieties of oranges (Citi·us over 4 years of 652 kg. sinensis [L.] Osb.), blond, blood and Cumulated yields/tree for each variety Navel, were tested (Table 2) over 4 years (trees aged 5 years in 1991, 6 years in 1992, 7 years in 1993 and canopy 8 years in 1994) are presented in Table 1. The canopy volume (23 m3, 3.2 m high x 3.7 m diameter) of orange trees in New The most productive cultivars were Caledonia is half that of trees grown Shambar and Marsh, with 750 kg and 727 kg cumulated 4-year yields, respec­ tively. These yields were almost threefold Table 2 higher than noted for cv Marsh under dry Orange varieties studied in New Caledonia. tropical climate conditions (MADEMBA-SY, Blond oranges Blood oranges Navel oranges 1989). Cadenera SRA 232 Double Fine SRA 259 Atwood SRA 157 These cultivars performed particularly Hamlin SRA 97 Sanguinelli SRA 243 Gilette SRA55 well in New Caledonia, apart from the Maltaise SRA 237 Navelate SRA 307 poor fruit colouring observed in some Pineapple SRA 142 Navelina SRA 306 varieties (Shambar, Redblush, Thom­ Sharnouti SRA 299 NewhaU Navel SRA 182 Valencia Late SRA 105 Washington SRA 141 pson).

Fru its, vol. 49 (5-6) / Special on Tropical Orchards • 365 • • MADEMBA-SY et al.

under wet tropical climate conditions Skin adhesiveness was high to average in (MADEMBA-SY & Corn , 1988). The trees the three types of oranges. Generally, the of all 14 varieties had a rounded habit. Navel varieties were easier to peel than the other varieties, except for cv Hamlin. ripening period The oranges generally had a yellow inter­ The orange trees flowered in September. nal and external orange fruit colour. Vety The ripening period was longer than early and early varieties (February-March) r that of grapef uits. Hence, cv Newhall had a less intense and even nonexistent had the shortest flowering-maturity period external fruit colour; full season varieties (5 months) and cv Valencia had the were definitely yellow, and end of the longest (1 0 months), with a mean of season, late and ve1y late varieties had a 6.5 months for all varieties. marked yellow external fruit colour. The The ripening periods for each variety appearance of anthocyanin pigments studied are given in Table 3. could be explained by the low tempera­ tures (less than 10-15'C) that occurred during the ripening period. Table 3 Ripening periods for the 14 orange varieties studied. production Very early Early Full season Very late The mean orange weight was 280 g, rang­ February March April July ing from 445 g for cv Gilette to 170 g for Newhall Atwood Cadenera Valencia Late cv Sanguinelli. Navel varieties had a ve1y Gilette Double Fine high mean fruit weight (376 g) relative Hamlin Maltaise to that of the blond and blood varieties Navellate Pineapple Washington Sanguinelli (213 g). Mean yields/tree for all varieties were 44 kg at 5 years, 79 kg at 6 years, 96 kg at fruit quality 7 years and 105 kg at 8 years, for a mean cumulated yield/tree over 4 years of Fruit quality could not be assessed in two 324 kg. varieties (Shamouti and Navelina) since they had not yet flowered. The results Cumulated yields/tree for each variety were thus based on a sample of 12 varie­ over 3 years (trees aged 6 years in 1992, ties. 7 years in 1993 and 8 years in 1994) are Navels were more oval shaped (ED/PD = given in Table 4. 0.91) than the other orange varieties. They were found to be thin skinned (4.5 mm). There were generally few seeds Table 4 in all varieties (n = 3), and the Navel Cumulated yields/tree for orange varieties were seedless, except for varieties over 3 years. cv Navelate. Variety 3-year cumulated The varieties differed in terms of the juice yield (kg/tree) content and soluble dry extract levels. Navel varieties had low juice (46%) solu­ Sanguinelli SRA 243 384 Pineapple SRA 142 ble dry extract (SDE = 8) and acidity 368 Hamlin SRA 97 309 (A = 1 g) levels. There was very little dif­ Atwood SRA 157 307 ference between blond and blood Washington SRA 141 300 oranges. They were found to be juicier Newhall SRA 182 289 (53%), sweeter (SDE = 8.35) and slightly Cadenera SRA 232 285 more aciduous (A = 1.05 g) than the Valencia Late SRA 105 244 Navel cultivars, thus providing them with Navelate SRA 307 231 more character. Overall, cv Hamlin was Gilette SRA 55 220 the juiciest (63%), sweetest (SDE = 8.9) Double Fine SRA 259 194 and least aciduous (A = 0.96 g) of all the Maltaise SRA 237 180 orange varieties studied.

366 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES e

Although cvs Sanguinelli and Pineapple were the most productive, they had the Table 5 lowest fruit weights. Navel varieties, how­ Mandarin varieties studied in New Caledonia. ever, produced high yields without any Clementines SRA 63 negative effect of fruit weight. Yields of (Citrns clementinci Hort ex Tan) SRA 64 cv Valencia Late in New Caledonia were SRA 85 higher than obtained under dry tropical SRA 92 climate conditions (175 kg). Satsumas Kawano SRA 167 (Citrns unshiu.Marc.) SRA 225 Saint Jean SRA 108 mandarins and hybrids Wase SRA 230 Thirty-one mandarin varieties were com­ Mediterranean mandarins Commune SRA ll8 (Citrus deliciosa Ten.) pared (Table 5). Large-fruit mandarins King of Siam SRA 273 canopy ( Citrus nobilis Lour.) SRA 166 Various mandarins Beauty SRA 261 Mandarin trees had a canopy volume of (Citms reticulata Blanco.) Carvalhal SRA 111 3 27 111 . The growth habit of the mandarin Dancy SRA 114 trees differed markedly: upright for cvs Ponkan SRA 234 Ponkan and Swatow, hanging for Satsuma Sanguine SRA 264 Swatow SRA 175 varieties and rounded for clementines. Hybrids (M. King x M. Commune) ripening period Kinnow SRA 26 Wilking SRA 112 All mandarins flowered in September. (M. King x Sarsuma) The flowering-maturity period ranged Kara SRA 165 r f om 5 to 9 months (mean 6.5 months) in (M. Dancy x grapefruit) the varieties studied. The classification for Orlando SRA 21 all varieties is given in Table 6. Minneola SRA 156 (Mandarin x orange) fruit quality Murcotr SRA 181 Ortanique SRA 110 The fruits were flat-shaped (ED/PD > 1), (Clementine x M. Dancy) except for cvs Carvalhal and Minneola Forrune S.RA 31 which were neck-shaped. The skins were (Clementine x M. Ponkan) fine (3.5 mm) to medium thick, and thick­ Fremont SRA 147 est for cv King of Siam (7 mm). The mean (Clementine x T. Orlando) number of seeds/fruit was increased by SRA 30 the presence of pollinating varieties Lee SRA 49 (Commune, Dancy). The Satsuma varie­ Nova SRA 158 Osceola SRA 48 ties had very few seeds (n = 3), or were seedless. Varieties with less than (Clementine x T. Minneola) 10 seeds/fruit were: Fortune, Fremont, Page SRA 159 Minneola, Orlando and Ortanique. Under these experimental conditions, clemen­ tines clearly had the most seeds (mean Table 6 28/fruit). Ripening periods for the different mandarin varieties studied. Mandarin varieties mainly differed in Very early Early Full season Late season Very late terms of juice content, acidity and soluble February March April May June dry extract levels. Mandarins (various, Carvalhal Sarsuma Saigon Clementine 63 Dancy Fortune large-fruit and Mediterranean varieties) Commune Satsuma Wase Beauty King of Siam Minneola had low juice content (37%) and high Satsuma Kowano Clementine 64 Fairchild Kinnow Satsuma St Jean Clementine 85 Kara acidity (A = 1.12 g) and soluble dry Clementine 92 Murcott extract (SDE = 8.6) levels. The clementine Fremont Orlando and Satsuma varieties were the juiciest of Lee Ortanique all the mandarins studied (42%). Nova Ponkan However, they were also the least sweet Osceola Sanguine (SDE = 7.66, as compared to 8.18) and Page Wilking aciduous (A = 1 g, as compared to Swatow

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 367 • • MADEMBA-SY et al.

1.07 g), which explains their lack of char­ productive variety group produced 2- to acter. The mandarin hybrids were much 4-fold more than levels obtained for crops juicier (49%), with as much as 59% juice grown in d1y tropical zones. content for cv Orlando and 61% for cvs Fortune and Lee. conclusion Low to high peel adhesiveness was The varietal improvement studies carried noted, i.e. it was low in clementine and out at the Pocquereux research station in Satsuma varieties and various large-fruit New Caledonia highlighted factors that and Mediterranean mandarin varieties, affect performance in about 50 citrus and average to high in all hybrids. varieties. Oranges and mandarins pro­ All mandarin varieties studied had duced high yields of top quality fruit remarkable internal and external fruit (internal and external) under the climatic colouring, much more intense than in conditions of the present study. oranges. Varieties that reached maturity in These preliminary results will be useful Februa1y (very early) or March (early) for guiding New Caledonian citrus grow­ had little or no colouring; they only tur­ ers in choosing varieties that are best ned yellow after degreening. From April adapted to the pedoclimatic conditions of on, colouring appeared in seasonal varie­ this archipelago. e ties, with intensely coloured fruits obtai­ ned from June to August. There were a few noteworthy exceptions: cv King of Siam and its hybrids Wilking and Kinnow were almost colourless, although they Table 7 reached maturity during the cold period. Cumulated yields/tree for mandarin In contrast to cv Fremont, cvs Beauty and varieties over 3 years. Page had a magnificent reddish-orange Variety 3-year cumulated colouring. yield (kg/tree) Minneola SRA 156 517 production Orlando SRA 21 477 The mean fruit weight for the vanet1es Sanguine SRA 264 444 studied was 177 g. Mandarin varieties had Dancy SRA 114 394 the lightest fruits (146 g), except for cv Satsuma Wase SRA 230 389 Beauty SRA 261 389 King (350 g), followed by clementine Fairchild SRA 30 388 (150 g) and Satsuma (158 g) varieties. The Kinnow SRA 26 373 hybrids (Photo 2) were generally the Nova SRA 158 337 heaviest (181 g), with high values obtai­ Ortanique SRA 110 325 ned for cvs Orlando (203 g), Kara (246 g), Commune SRA 118 313 Ortanique (288 g) and Minneola (297 g). Satsuma Kowano SRA 167 311 Clementine SRA 64 306 Mean yields/tree were 46 kg at 5 years, Clementine SRA 63 303 75 kg at 6 years, 93 kg at 7 years and Clementine SRA 85 302 123 kg at 8 years, for a mean cumulated Clementine SRA 92 298 Satsuma Saigon SRA 225 294 yield/tree over 4 years of 337 kg. Page SRA 159 281 Cumulated yields/tree for each variety Fortune SRA 31 281 over 3 years (trees aged 6 years in 1992, Kara SRA 165 280 7 years in 1993 and 8 years in 1994) are Satsuma St Jean SRA 108 253 given in Table 7. Carvalhal SRA 111 246 Osceola SRA 48 238 The least productive varieties (cumulated Wilking SRA 112 235 yields of < 220 kg for 1992-93-94 Swatow SRA 175 218 harvests) were Swatow, Lee, King of Lee SRA 49 201 Siam, Malvasio, Murcott, Fremont and King of Siam SRA 166 186 Ponkan. The most productive varieties Murcott SRA 181 186 Malvasio SRA 163 171 (cumulated yields of > 350 kg for 1992- Fremont SRA 147 170 93-94 harvests) were Minneola, Orlando, Ponkan SRA 234 164 Sanguine, Dancy, Satsuma Wase, Beauty, King of Siam SRA 273 145 Fairchild and Kinnow. Overall, this

368 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES •

Photo 2 Mandarin Page SRA 159.

references

GODEFROY J., 1990. Etude agropedologique de la station fruitiere de Pocquereux (Nouvelle-Caledonie). Mont­ pellier, France, IRFA-CIRAD, 132 p. (internal document).

GUILLAUMIN A., 1952. Les arbres fruitiers en Nouvelle-Caledonie. Fruits 7 (4) : 147-148.

MADEMBA-SY F. , 1989. Comportement de six varietes d'agrumes dans la zone des Niayes au Senegal. Fruits 44 (4) : 205-213.

MADEMBA-SY F .• cornNR .• 1988. Comportement de trois varietes d'agrumes sur plusieurs porte-greffe en climat tropical humide. Montpellier, France, Annual meeting of IRFA, document n"17.

ORSTOM, 1989. Atlas de Nouvelle-Caledonie. Nouvelle­ Caledonie : Hachette Caledonie, edition du Cagou, 91 p.

PRALORAN J.C., 1971. La production fruitiere en Nouvelle-Caledonie. Paris, France, Mission report, 80 p. • •••

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 369 • • MADEMBA-SY et al.

appendices

Appendix 1 Traits measured on the citrus Flowering date: COll'esponding to the Citrus phenological E stage (open flower). varieties studied. Maturity date: determined according to the city extract/acid ratio, which was set at 7.5 for oranges and mandarins and 7 for grapefruit.

Flowering-maturity period: from the E stage (open flower) to maturity as determined by the dry extract/acid ratio. Fruit weight: mean (g). Yield/tree: mean yield from four trees. Polar diameter (PD): diameter measured from the I a e of the peduncle to the apex (mm). Equatorial diameter (ED): fruit diameter (mm). Peel thickness: measured from the epicarp to the mesocarp (mm). Juice content (SDE): percentage juice relative to the fruit weight. Soluble sugars: dry extract measured by refractomeuy ("Brix). Acidity (A): citric acid content (g) after neutralization with a sodium base (NaOH N/ 10).

Appendix 2 Citrus phenological stage.

A Green bud B White bud C Round bud D Long bud E Open flower F fa ll Style fa ll B C G D E H Fruit-set - - Wa lnut-size ------CJ J Fruit swelling 1 cm

l \ F G H I J

Appendix 3 Citrus ripening periods in New Caledonia. Jan. Feb. March Apr. May June July Aug. Sept. Oct. Nov. Dec.

VE VE = very early -- E E early -- FS FS full season LS LS late season -- L L late -- --VL VL = ve1y late

370 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards Genetic Resources of Mangos in Cote d'Ivoire

T. GOGUEY ••••••••••••••••Ill ••••••••••••••••••••••••••••• IDEFOR/DFA A germplasm repository of the main mango varieties exported BP 856, Korhogo Cote d'Ivoire worldwide has been set up in northern Cote d'Ivoire. The genetic variability in this collection could be tapped for varietal breeding purposes once all accessions are characterized and their cropping potentials assessed. Fruits, vol. 49, n°5-6 •••••••••••••••••••••••••••••••••••••••••••••• p. 371-375 (English) p. 461-463 (French)

the Lataha germplasm all of the 30 or so most common export repository varieties. Ultimately, a germplasm reposito,y varietal inventory should contain a wide variety of genetic material that can be used in varietal Interesting Florida mango varieties (, improvement programmes. It also should , , Smith, , Ruby and be regularly updated and evaluated. Valencia) were introduced at the Lataha reposito1y immediately after the station was founded in the Korhogo region of accession origins northern Cote d'Ivoire in 1981. This is still The 137 mango varieties in the Lataha acknowledged as being a valid germ­ repository were collected in nine different plasm pool. An experimental orchard was countries: Cameroon, Canary Islands, set up in 1982 to test their performance Cote d'Ivoire, Egypt, Guadeloupe, and that of cv Amelie (also known as cv Guinea, Indonesia, Mali and Martinique. Gouverneur and cv Mangue Greffee), These accessions actually originate from which is common throughout western Florida, the West Indies, No1th and South Africa. At the same time, several other America, Africa and especially Asia. varieties were introduced, including cvs , Springfield, Late, Tommy The poor representation of accessions of Atkins, Broocks, Early-Gold, Haden, Asian origin could be explained by the Amelioree du Cameroun, Eldon and prevalence of some bacterial diseases . (Xanthomonas campestris) and the unat­ A project to expand the initial collection tractive colour of mangos from this with other widely-marketed varieties was region. However, the genetic variability undertaken in 1989. There are thus now offered by these varieties, which differ 137 , M. laurina and phenotypically from the Florida varieties, M. pelipisan varieties in the repository; could be of considerable interest for this plant material is listed in Table 1. varietal breeding programmes. Some The collection is still far from being Asian varieties produce high yields which exhaustive since there are 1 100 known are consumed domestically in India M. indica varieties and 61 Mang[fera sp. (9-10 Mt); few of these are represented in worldwide; nevertheless it does include the Lataha reposit01y.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 371 Table 1 List of mango varieties in the Lataha reposito1y in northern Cote d'Ivoire.

Variety Collection location Variety Collection location

Adams Canaries via Cameroon lfac 1 Cameroon Aeromanis Mali Jr\'.\rin Cote d·Ivoire (Azaguie) Ah Pingh Canaries Jacqueline Mali Alphonse Mali Je 4 Guinea Alphonse de Goa Mali Jose Cameroon Alphonse Hawai Cameroon Julie Lataha Cote d'Ivoire Alphonse Paheri Mali Keitt Cote d'Ivoire Ameeri Guinea Kensington Mali Amelie Core d'Ivoire Kensington Cameroon Amelie Guinea Mali Kent Cote d'Ivoire Amelioree du Cameroon Cote d·Ivoire Linzolo Mali Canaries Linzolo Mali Auguste Cameroon Guinea Bedami Rouge Mali Lippens Cote d'Ivoire Bedami Vert Mali Longo Diego Mali Beverly Cote d'lvoire Mahroka Cameroon Big Yellow Can�1ries Mabrouck Cote d'Ivoire Black Guinea Maison Rouge Reunion Bombay Mali /Vlalembe Mali Bombay Green Canaries /Vian Guadeloupe Boy Toy 15 Canaries via Cameroon Mangotine Mali Broocks Cote d'Ivoire Mangot Veit Mali Camayenne Camero n Ma1tin Guinea Cambodiana Cameroon Maya Mali Cameroon Mechouang Guinea Cecile Mali Miami Late Cote d'Ivoire Christian Mali M. /a11rina Indonesia Coq's Hall Mali .M. pelipisan Guadeloupe Crazou Mali Mulgo Round Guinea Cuisse Mali Nimrod Mali Dabsha Drahnet Mali Noura .Mali Davi:; Haden Mali Oclet.te Mali Diego Mali Canaries via Cameroon Divine Cote d'Ivoire Paheri Guinea Dixon Mali Palmer Core d'Ivoire Djibelor Senegal via Cameroon Paris Mali Djibelor Casamance Mauritania via Cameroon Passy Harive Mauritania via Cameroon D'or Mali Peche Core d·1voire Early Gold Cote d·1voire Peter Passancl Mali Guinea Petit Greew Guinea Egypte a Egypt Philipino Canaries Egypte b Egypt Pico Mali Egypte c Egypt Pirie Canaries Egypte cl Egypt Pomme Cameroon Egypte e Egypt Pope Canaries Egypte f Egypt Rachel Mali Eldon (1) Cote d'lvoiJ·e Romania Mali Eldon (2) Cote d·rvoire Ruby Cote d'Ivoire (Azaguie) Eldon Ferke Cote d'Ivoire Saber saber Eugenie Mali Sabot Mali MaU Sabre Cote d'Ivoire Ferke Cote d·Ivoire Sabre Cote d'Ivoire Canaries via Cameroon Sacabi Mali Frai sinette Cote d'Ivoire Sans Pareil Mali Francis Mali Cote d'Ivoire Gabriel Guinea Smith Cote d'Ivoire Galerie Cote d'Ivoire Souclan 2 Cameroon Gedong Mali Springfield Cote d'Ivoire Glazier Mali Sybil Mali Glazier Cameroon Taymour Canaries Canaries Tolbert Canaries via Cameroon Golek Mali Mali Gomera a Canaries via Cameroon Tommy Atkins Cameroon Gomera 2 Canaries via Cameroon Ton=y Atkins Cote d'Ivoire Gomera 3 Canaries via Cameroon Valencia Cote d'Ivoire Gordon Guinea Canaries Grosse Rouge Mauritania via Cameroon Canaries via Cameroon Haden Core d'Ivoire Whitney Cameroon Haden x Carabao Mali Wooten Guinea Heart Mali Xoi-Cat-Mytho Mali Hinclibi Sinara Mali Zill Cote d'Ivoire Hybrides Cote d'Ivoire 9c Mali • GENETICS / GENETIC RESOURCES e

varietal characterization Table 2 standard varietal descri ption sheets Characters involved in the floral biology of mango trees and which provide information on panicles and flowers. This supplements the Mango varieties in the Lataha repository criteria considered in the IPGRl character description sheet. are assessed according to descriptors published by the International Board for Number of flowers Plant Genetic Resources (JBPGR, now Flower colour (qualitative) IPGRI). The problem is that the Indian Distribution of flowers on the inflorescence apex (%), base (%) team that drew up this report Sex ratio per inflorescence pa11 apex (%), base (%) (ANONYMOUS, 1989) determined the Pollen life (shott or average or long) descriptors only on the basis of varieties Pollen germination (short or average or long) -pistil length ratio grown in India. Moreover, they did not Ovary condition (description) take into account some traits (e.g. colour­ ing and fruit size) that are unstable in cer­ tain environments. Despite these minor theoretical problems, use of the IBPGR varietal description varietal improvement sheets should become standard practice in order to homogenize the characteriza­ breeding objectives tion of mango varieties worldwide. The basic description could be supplemented Breeding objectives are set according to with other details in the light of new consumers' and producers' needs, which research fi ndings. A chapter on the floral are not always identical. This means that biology of the tree, based on studies of many different criteria have to be taken the mango collection at Lataha, has there­ into account, especially those having an fore been added (Table 2); data on pani­ impact on quality. cles and flowers can thus be recorded. The mango varieties Arnelie, Zill, Kent, Fruit quality, a critical parameter for con­ Keitt, Valencia, Palmer, Early Gold and sumers, is obtained by controlling various Broocks have been described since 1990 constraints: grade, organoleptic qualities using the standard data sheets; cvs and appearance of the mango fruit, and Haden, Smith, Ruby, Miami Late and also crop protection, shipping and storage Beverly are now being described. qualities. The essential criteria for produc­ Systematic use of these varietal descrip­ ers are earliness of producing, yield, har­ tion sheets could highlight traits to be vest regularity and length of this period. modified in breeding programmes. Work involved in breeding programmes based on such criteria requires consider­ use of the data able labour, space and especially time. A Descriptions of different commonly mango tree is only considered to be fully grown mango varieties should facilitate mature after 8- 10 years growth; the tree characterization, between-variety compar­ has to be another 3-5 years older before isons, genetic distance evaluations, and these production criteria can be evalu­ possibly parentage determinations. In ated. Multilocation performance studies turn, genetic analyses could be conducted should also be carried out in different on the basis of calculations of correlations regions to determine the often consider­ between variables, heritability, genetic able influence of the environment (cli­ variability and environmental factors mate, soil, diseases, parasites, etc.) on the (PERRIER, pers. comm.). phenotypic expression of certain mango The observations are therefore both: characters. - qualitative, with a theoretical tree repre­ sentative of the variety taken as the basic The results of a prelimina1y quantita­ unit; tive/qualitative study of this type with - and quantitative, with data pertaining to some mango varieties are given in one or more trees at a given site. Table 3.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 373 • e GOGUEY

varietal breeding cv Lippens, cvs Keitt and Haden from cv , and cv Kent from cv Broocks. controlled hybridization About a hundred pits from six monoem­ Only two countries have a structured bryonic varieties (cvs Kent, Valencia, varietal breeding programme, i.e. India Broocks, Zill, Amelie and Early Gold) (Sharma, 1987) and South Africa (TOMER et have been sown at the Lataha station in al., 1988). Cote d'Ivoire. Following two years of pro­ At the Lataha station (in northern Cote duction, the traits of a few trees have d'Ivoire), a controlled hybridization pro­ been found to clearly differ from those of gramme began in 1989 using parents that their mother-trees. The phenotypes of were selected for specific characters of fruit from these variant trees (grade, agronomic interest. Five fruits have been colour, shape) were assessed visually. obtained (two ripened) since 1991 using a Further in-depth studies should be carried hybridization technique adapted to local out in the coming years to assess the conditions (GOGUEY, 1991). The two pits, agronomic potentials of these new varie­ from fruits obtained by gamete fusion ties. If they are found to have interesting with cv Amelie (female parent) and cv quality and yield features, these new gen­ Early Gold, were successfully sown. otypes could be vegetatively propagated Scions were cut from the resulting plant­ from adult mother-trees. lets after 1-month growth; five trees have now been grown from each of these hybrids at the Lataha station. These trees conclusion have not yet flowered or fruited; the long juvenile phase in mango tree develop­ The Lataha mango germplasm reposito1y ment is a hurdle that will have to be over­ in northern Cote d'Ivoire is of interest for come by suitable techniques (single-axis evaluating the performances of a wide growth management, arcure training, range of mango varieties in the Sudano­ etc.). Sahelian climatic environment of western Africa; it is also a useful pool for varietal Hybridization of mango is an especially breeding. The reposito,y will therefore be long process requiring a great deal of pre­ a natural asset for promoting develop­ cision because of the size of the flowers ment of the mango sector in the coming (4-6 mm), self-fertilization (which auto­ years. e matically emasculates hermaphrodite flowers) and low fertilization rates in this fruit tree. In addition, the resulting hybrids have to be monitored for long periods to assess their agronomic qual­ ities. references open pollination

The so-called random hybridization tech­ ANONYMOUS, 1989. nique is much easier to set up. It involves Descriptors for mango. Rome, Italy, IBPGR, sowing a number of pits obtained from 22 p. fruit of monoembryonic varieties. The sexually-obtained emb1yo should give GOGUEY T., 1991. rise to a tree producing fruit that differs Collections etablies en Afrique de l'Ouest et sur le reseau IRFA. Korhogo, Cote-d'Ivoire, markedly from that produced by the Annual meeting of IRFA n"66, 21 p. mother-tree. Many Florida varieties have been obtained in this way, and most of SHARMA D.K., 1987. the mangos marketed worldwide are of Mango breeding. In: Acta Horticulturae, num­ this ongm. Varieties Tommy Atkins, ber 196 : 61-67. Smith, Springield, Valencia, Zill, Eldon, TOMER E., SNYMAN J.C., MULLINS P.D.F., 1988. Osteen and Glenn, for instance, are deri­ Mango selection at the CSFRI and recom­ ved from cv Haden seed; cv Broocks is mendations for a breeding programme. derived from cv Sandersha, cv Irwin from Nelspruit, South Africa, CSFRI, 4 p.

374 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards Table 3 Internal and external features of a few mango varieties.

Fruit Pulp Pit

Variety Harvest Long. Width Thick. Weight Shape Colour Appear Taste Fibres F. long. Pulp Pulp Peel Long. Weight Fibres F. long. Storage (cm) (cm) (cm) (g) (*) (*) (*) (•) (%) waste thick. (cm) (g) (*) (*) life (mm) (d)

Beverly Season 11.1 8.9 8.8 487 Round "Reel" 7 6 4 7 70.0 2.30 0.48 8.5 51 5 9.0 30

Miami Late Late 12.2 10.1 9.3 625 Round "Mixed" 5 4 2 5 60.3 1.52 0.44 9.5 49 5 2.6 30

Ruby 1/2 early 9.3 6.1 5.6 189 Long "Red" 7 7 3 5 72.4 2.63 0.52 7.7 22 7 7.7 20

Broocks Late 11.3 7.7 7.2 315 Oblong "Green-yellow" 2 3 7 5 55.5 1.25 0.52 9.3 53 6 7.5 35

Keitt Late 12.5 9.7 7.8 465 Round "Mixed" 6 7 6 6 66.0 1.94 0.72 10.7 39 6 8.4 35

Valencia Season 13.9 67.8 6.9 415 T. oblong "Mixed" 6 7 7 7 67.5 2.10 0.53 11.7 43 6 7.5 25

Zill Early 91 7.3 6.6 345 Round ''Red" 7 8 7 7 62.9 1.70 0.71 6.5 30 5 3.9 20

Kent Early 12.2 10.6 9.6 470 Round "Red" 8 8 1 4 72.5 2.64 0.47 9.9 48 4 4.6 30

Palmer Season 14.8 8.5 7.8 483 Oblong "Purple" 6 6 8 8 45.6 0.84 0.50 12.6 86 5 4.2 30

Early Gold 1/2 early 12.8 8.1 7.1 365 Oblong "Red" 8 7 4 7 63.1 1.71 0.30 9.0 41 5 4.3 25

Amelie Early 10.4 8.4 7.2 371 Round "Orange" 5 8 3 3 75.2 3.03 0.66 7.3 17 4 2.3 30

(•) Data correspond to a variety rank (scale from 1 to 20). Management of Litchi Genetic Resources in Reunion

F. NORMAND •••••••••••••••••••••••••••••••••••••••••••••• J. 80UFFIN A litchi germplasm repository was set up in Reunion to diversify CIRAD-FLHOR BP 180 litchi crops and provide a basis for breeding programmes. 97 455 Saint-Pierre cedex Fruit and floral descriptors were determined to characterize Reunion litchi varieties. Fruits, vol. 49, n°5-6 •••••••••••••••••••••••••••••••••••••••••••••• p. 376-382 (English) p. 464-468 (French) introduction cv Kwa·i-Mi. The following features are being sought: Litchis Sann.) were first (Litchi chinensis - lengthening the production period, brought to Reunion (Indian Ocean, 21" 20' - improving resistance to cyclonic winds, latitude S, 55"25' longitude E) by Pierre - reducing biennial bearing, POIVRE in 1779. They were cropped on 111 - increasing fruit quality and storage the island throughout the 19 centu1y; potential. there are now more than 1 OOO ha of litchi orchards, making it the most widely This repository provides a litchi pool that grown fruit crop in Reunion. is utilized for varietal breeding pro­ grammes whose results have a direct These orchards are mainly cropped with impact in the field. Varietal descriptions cvs Kwa·i-Mi and Ta·i-So, which produce are also under way with this plant mate­ excellent-flavoured litchis that are suitable rial. for both domestic and export markets. There are also some so-called "cv Litchi Toupie" trees that produce a spinning-top setting up the litchi repository shaped fruit with an aborted pit. accession origins The shortness of the production period The first mass introduction of litchi varie­ could be explained by the lack of diver­ ties took place in 1985 with 26 different sity in litchis of Reunion, mainly due to varieties that had been imported from the conventional use of the air marcotting Australia, India and the Seychelles. technique for propagating this fruit. Other varieties were introduced after Litchis are harvested from late November being collected in various countries, i.e. to mid-Janua1y because the climate differs Bali, Hawaii, Mauritius, New Caledonia on the windward and leeward sides of and Thailand. the island. In addition, there are several drawbacks concerning cv Kwa'i-Mi, i.e. These litchi varieties were generally intro­ the fruit does not stand up well to cyclo­ duced in the form of suckers and scions, nic winds, it shows marked biennial bear­ and less commonly as seeds. ing and high vigour. Eleven different varieties producing fruit that differ from that of cv Kwa'i-Mi were CIRAD-FLHOR has set up a litchi germ­ obtained in surveys conducted in Reunion plasm reposit01y at different locations in (NORMAND, 1990a). Reunion with the aim of broadening the varietal base and minimizing drawbacks Table 1 provides a list of the different that arise as a result of monocropping varieties and their origins.

376 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES e

locations of litchi introductions Table 1 The first litchi introductions in Reunion List and origins of litchi varieties in the CIRAD-FLHOR repository took place in 1986 at Bassin-Plat on a in Reunion. 150 m elevation site in the southwestern part of the island. There were further Origin Variety Origin introductions on this plot in 1989, with varieties of Hawaiian origin, and in 1990, Kwa1-Mi Reunion Borsworth Australia with varieties collected in a local survey. Gros fruit (2) Reunion Salathiel Australia Lisse (2) Reunion Thailand Australia In 1990, the collection was partly dupli­ Lisse coeur Reunion Wai Chee Australia cated at Bassin-Martin (Reunion), on a Piquant (2) Reunion Brewster Australia and Hawaii• Tardif 300 m elevation site in the same part of Reunion Groff Australia and Hawaii" Toupie (3) Reunion Haak Ip Australia abd Hawaii" the island. Violet Reunion In 1993, a few selected varieties were Litchi X Reunion Takafuji Hawaii Litchi Y Reunion Kwa1 Mi Hawaii Hawaii planted on a smallholder's farm, as part Kawai (Hawaii)? of a CIRAD research programme funded Al Seychelles by ODEADOM to investigate litchi crop­ B3 Seychelles Emperor Thailand ping under real field conditions. This site C4 Seychelles is at 300 m elevation in the southeastern Dl Seychelles Bali Planteur Bali part of the island (Sainte Rose). D2 Seychelles E7 Seychelles Mauritius (Mauritius)? A number of varieties planted in 1986 G2 Seychelles Kwat Mi Maurice (Mauritius)? have been flowering and producing since PDM Seychelles Litchi Nlle-Caled. New Caledonia 1988. Flowering was also noted in 1992 Calcutta India on trees planted in 1990. Dehradum India Kwat Mi Nialo Muzzafapur India Kwa'i May Pink Saharampur India Soueytong varietal breeding Rose Scented (India)? Hueng Lai Seedless Late (India)? selection criteria Bengal (India)? (n) Number of varieties • Variety introduced as seed There are only one, two or three trees representing each variety in the reposi­ tory, thus fruit quality and earliness of harvest were the main characters that - adhesiveness of the shell to the aril, and could be selected. Some agronomic fea­ the aril to the pit, tures could also be taken into considera­ - shell and aril colour, tion, e.g. earliness of production, tree vig­ - juice acidity and dry extract, our, graft compatibility on cv Kwa'i-Mi, - aril moisture content. shortness of the fruiting period and pro­ ductivity. However, the degree of reliabil­ Analyses were conducted with 11 varie­ ity of the results could not be assessed ties in 1989, including Kwa'i-Mi (NORMAND, because of the low number of trees/ 1990b), and 17 varieties in 1991 (BERTIN, variety tested. Detailed and more statisti­ 1992). cally-valid analyses of these characters will thus be carried out in the future in harvest dates experimental orchards cropped with improved varieties. Earliness of the harvest is an essential consideration for evaluating the spread of fruit quality the production period. Litchi producers seek early varieties to be able to supply Fruit quality was defined by the following criteria: world markets before competing litchi­ - mean fruit weight, producing countries in the region, i.e. - fruit size, Madagascar and Mauritius. Late varieties - fruit flavour, are a drawback because they produce - percentage of different fruit components during the cyclone period, which can (shell, aril, pit), lead to considerable crop losses.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 377 • • NORMAND and BOUFFIN

analysis of tested criteria - cv Dehradum, cv PDM (Seychelles) and cv Haak Ip: ve1y high quality varieties The percentages of aril and pit, the main with high fruit weights, low pit and high factors studied in the two analyses, were aril percentages, and very flavourful fruit; similar in 1989 and 1991 (Table 2). These - cv B3 (Seychelles) and cv Bali-Planteur: important qualitative parameters did not with an early harvest date, average acidity seem to va1y between years at a given and aril percentage; site. - cvs Muzzafapur, Rose Scented and Thailand: with low aril and high pit per­ centages, and low fruit weight. Table 2 Aril and pit percentages for the different varieties in 1989 and 1991. All other varieties were shared between these groups. Percentage aril Percentage pit Variety 1989 (*) 1991 (••) 1989 (•) 1991 (**) The results revealed:

Kwa'i-Mi 75.2 70.0 10.3 12.8 • varieties with interesting characters: Kwa'i May Pink 64.5 67.5 11.7 11.0 Dehradum, Haak Ip, Kwa·i-Mi, Kwai May Thailand 62.4 56.5 10.1 25.5 Pink, Soueytong, Mauritius, Bali-Planteur Calcutta 62.7 62.4 15 6 15.6 and three varieties from the Seychelles Bengal 49 7 56.2 18.7 20.4 (B3, D2 and PDM). The local variety D2 57.5 60.4 14.1 14.6 Kwai-Mi was high in this classification. Brewster 52.9 56.7 18.1 21.3 Two varieties were markedly better than C-) Normand, 1990b (••) Bertin, 1992 the others, i.e. PDM (Seychelles) and Haak Ip, especially in terms of frnit fla­ vour; The harvest dates for varieties that yiel­ ded in 1988 and 1989 (NORMAND, 1990b) • varieties with characters of little interest and 1991 (BERTIN, 1992) in the Bassin-Plat for cropping in Reunion: Tha'iland, collection indicated that: Brewster, Saharampur, Calcutta, Muzza­ fapur, Rose Scented and Bengal. Note that • most of the varieties produced during several of these were of Indian origin. the first half of December at Bassin-Plat, i.e. almost the same time as Kwa·i-Mi, The best varieties, i.e. Haak Ip, PDM which meant a harvest-date gain of only (Seychelles), Dehradum and Kwa·i May about 10 days over Kwai-Mi (Seychelles Pink, will be planted in experimental variety D2, cvs Dehradum, Mauritius and orchards for varietal comparisons with Kwai May Pink); cv Kwa·i-Mi. Production processes of new varieties in the collection will be moni­ • only a few varieties were found to be tored to add to the present results and very early (cv E7 (Seychelles) - mid­ determine a set of interesting varieties to November harvest; cv B3 (Seychelles) and improve litchi cropping in Reunion. cv Bali-Planteur - late November harvest) or very late producers (cv Calcutta - early January harvest). varietal characterization • harvest dates varied substantially bet­ historical background ween years, but the spread of the harvest period was not modified for extreme Litchis are originally from southeastern varieties (i.e. very early or late), and only China. The qualities of this fruit have slightly modified for other varieties. This been renowned for centuries in this coun­ variable is therefore a valid selection cri­ try and the first descriptions were Chinese terion. (LIANG, 1981). Chinese names for litchi varieties are very imaginative, but often breeding interesting varieties ignored or changed when they are intro­ duced outside of China: the Chinese Multivariate analyses involving the main name is sometimes translated, for instance quality variables highlighted groups of cv Hei Ye, means "black leaf", which is its varieties with similar characteristics varietal designation in USA; sometimes (BERTIN, 1992), these were: the original phonetics are kept but the

378 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES e

spelling is changed, e.g. cv Haak Ip - fruit quality descriptors which are a comes from the Chinese "Hei Ye", and prime consideration for varietal breeding; cv Tai So is derived from "Da Zao"; some­ these were described above; times the accession is named after the - taxonomic descriptors; to unify the person who first introduced it, such as descriptions (shapes of fruit, segments cvs Groff and Brewster. and protuberances), these descriptors are generally in line with the conditions set Moreover, litchis were long ago intro­ out by MENZEL & SIMPSON (1986). duced in India; this countiy is probably However, some other characters have the second centre where the species been added, including the shell suture diversified. Many litchi varieties have line and the pit. Indian names and it is often quite difficult to determine whether a variety is of These data were recorded for litchis col­ Indian or Chinese origin. lected in a local survey (NORMAND, 1990a), and for varieties that began fruiting in There is indeed a great deal of confusion 1988 (NORMAND, 1990b; BERTIN, 1992). concerning litchi variety names and a detailed morphological description of floral descriptors each variety is necessa1y to supplement studies on the origins of the present NORMAND et al. 0990) defined 22 quanti­ variety names. tative and qualitative floral descriptors, which are divided into three groups: Characters that are most often used to - general flower characters, irrespective of identify litchi varieties pertain to fruit the flower type (male or female), descriptions (BATIEN, 1984; ANONYMOUS, 1985; MENZEL and SIMPSON, 1986). They can be associated with vegetative charac­ Table 3 ters to describe the most commonly mar­ List of fruit descriptors used to characterize litchi varieties. keted varieties. SINGH and SINGH (1954) introduced inflorescence-related charac­ Organ Descriptor ters in their descriptions of Indian litchis. However, current descriptions are incom­ Whole fruit Shape of fruit, shoulders1 , apex2 plete and deal chiefly with production; Dimensions: length, diameter 13, diameter 24 (mm) they provide little information to clarify Mean fruit weight (g) Percentage of twinned fruit5 the taxonomy. Suture line on the fruit: marked, barely visible, undetectable Percentage of different components: shell, aril, pit research under way in Reunion Shell External colour at maturity CIRAD-FLHOR studies conducted in Internal colour at maturity Reunion since 1989 have focused on fruit Thickness (mm) 2 characters and, more originally, floral Protuberances: general shape , length and width at the base, height, number/cm2 characters. There are several reasons for this choice: Aril and juice Shell-aril adhesion: strong, average, weak - flowers are commonly used in determi­ Aril-pit adhesion: strong, average, weak nation keys, Colour - floral characters are often the most Soluble dry extract at maturity (E) in °Brb: stable types in various environments Acidity at maturity (A) in meq/100 ml E/A (HILU, 1989), Flavour - no detailed studies have been under­ Moisture content (% fresh weight) taken to date on litchi flowers, - litchi varietal characterizations could be Pit Pit description: shape, colour performed from the flowering stage, Dimensions: height, diameter 13, diameter z4 which is earlier than the harvest stage. Percentage aborted pits fruit descriptors (1) swelling at the peduncle (2) see MENZEL and SIMPSON (1986) The litchi fruit descriptors defined by (3) diameter along the suture line axis NORMAND (1990b) can be divided into two ( 4) diameter perpendicular to diameter 1 (5) fruit with both ovaries developed, thus forming a double-joined fruit groups (Table 3):

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 379 • • NORMAND and BOUFFIN

- characters and conditions specific to stability results. It seems that quantified male flowers, descriptors are generally only useful for - characters and conditions specific to comparisons of extreme values (COMMON, fe male flowers. 1993). The flower sampling and monitoring LITAID has a no-answer function which techniques for measuring quantitative enables identifications with only a few characters are described by NORMAND et descriptors. al. (1990). The program includes a module to calcu­ These data were recorded for 11 varieties late phenotypic distances; this is the in the Bassin-Plat collection in 1989 mean distance between two individuals, (NORMAND et al., 1990) and for 36 varie­ for all descriptors, weighted by the asso­ ties, including 23 from the Bassin-Plat col­ ciated error fu nctions. This distance is not lection and 13 from the young Bassin­ a real distance in mathematical terms, Martin collection, in 1992 (COMMON, because at zero there is no probability in 1993) the matrix. The distance only indicates the extent of convergence or divergence data management of the studied character for two individu­ The quantity of fruit and flower data als within the reliability limits of these recorded is ve1y important, and they characters. should be computerized for efficient man­ agement. The LITAID software program was developed by the CIRAD-FLHOR varietal choices biometry service (NORMAND et al., 1990). It validity of some floral characters is a derivative of the MUSAID program that was designed to manage morphotax­ The initial flower results highlighted some onomic data and provide an identification intervarietal diversity in the floral charac­ aid for banana (PERRIER & TEZENAS DU ters (Photo 1), which led to the develop­ MONTCEL, 1988). ment of a determination key for the 11 varieties studied (NORMAND et al., LITAID has the same two main functions 1990). The first two dichotomies of this as MUSAID: key are based on two stable characters, r - f uit and flower data can be managed; stamen number and colour of the necta­ - it is an identification aid which enables riferous disk, under clear-cut conditions; the operator to match an unknown plant they are also independent of the flower with an already identified and described type and easy to monitor. variety, with a reasonable margin of error. Following the second set of observations The operating principle is the same as (COMMON, 1993), the validity of the that of MUSAID. The unique aspect is that descriptors was assessed by studying each descriptor is provided with an error flowers of certain varieties grown at the matrix, weighting it relative to its reliabil­ same site (Bassin-Plat) at 3-year intervals, ity. A descriptor with two distinct stable and at two different sites (Bassin-Plat and conditions will thus have a low error Bassin-Martin) during the same year. fu nction; in contrast, when two or more Comparisons of descriptions of the same conditions are difficult to distinguish or variety for different years and different are not very stable, the error fu nction will sites revealed the high stability of two be higher. characters, i.e. the colour of the nectarif­ Determining the error matrix for each erous disk and the anther shape. Other descriptor is a critical step because it characters were found to be relatively clearly specifies the degree of reliability stable, i.e. anther length on male flowers,

and quality of identifications for unknown stamen number and calyx colour plants. (COMMON, 1993). After the second set of observations in The conditions associated with each vari­ 1992, these error matrices were refined as able were also defined in further detail on a function of the descriptor reliability and the basis of the latter results.

380 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES •

determining groups Three groups were determined on the basis of the calculated phenotypic dis­ tances between 22 varieties from the Bassin-Plat collection (COMMON, 1993). Two very close groups, which differ in terms of the colour of the nectariferous disk, are formed by Kwa·i-Mi type varie­ ties: one includes cvs Kwa'i-Mi, Kwa·i-Mi Maurice, Kwa·i-Mi Nialo, Takafuji, B3 (Seychelles), and three locally-collected varieties (cvs Gros Fruits 11·2, Litchi Tardif and Litchi Piquant); the other includes cvs Mauritius, PDM (Seychelles), Haak Ip and Dehradum. The third group comprises three varieties of Indian origin, cvs Calcutta, Bengal and Rose Scented. Although incomplete, since only the floral characters were taken into account, the present results are of interest for the study of genetic resource diversity in litchi. Photo 1. Moreover, varieties that were grouped on Female flowers for cv Kwai'.Mi the basis of their highly convergent floral (left) and a variety from the Seychelles (right). characters also had similar fruit charac­ r ters. The fruit produced by varieties Kwa·i­ Varietal studies using f uit and floral Note the marked morphological Mi Maurice, Kwa·i-Mi Nialo, B3 descriptors are currently under way. This differences in the shape ° of the calyx, protuberance (Seychelles), Gros Fruits 11 2 and Litchi work is facilitated by the LITAID software program that was designed to manage of the nectariferous disk, Tardif, classified in the first cv Kwa1·-Mi length of the stamen filaments, related group, was similar to that pro­ morphotaxonomic data and to serve as an identification aid. It now handles flower length and shape of the duced by cv Kwa'i-Mi and only differed in stigma. and fruit data for several varieties. Groups their sizes and mean fruit weights. The can be determined on the basis of the fruit of cv Takafuji has not yet been stud­ stable reliable floral descriptors. ied. The locally-collected cv Litchi Piquant was the only one of this group to differ One of the main objectives now is to markedly from cv Kwa'i-Mi in terms of enhance the characterizations with new fruit and leaf characters. phenotypic, vegetative and inflorescence descriptors, and with biochemical mark­ Similarly, in the second group, cvs PDM ers. (Seychelles), Haak Ip and Dehradum have already been grouped according to The current system could be optimized fruit-related criteria considered for varietal and utilized by: breeding purposes. - reducing the number of descriptors used in flower studies, - analysing the results already obtained on fruit and on fruit/flower combinations, prospects - comparing local results with those obtained elsewhere in the world, At various locations in Reunion, CIRAD­ - building up the reference data bases. FLHOR has a large reposito1y of litchi varieties of different origins. The first New litchi varieties should be rationally accessions have already been bred to pro­ introduced in Reunion to enable continu­ duce four hybrids of the same quality as ous selection of well-adapted varieties the local variety Kwai-Mi, but they can be and broaden the scope of varietal charac­ harvested slightly earlier. terization research. •

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 381 • • NORMAND and BouFFIN

references Description sheet section : Litchi ANONYMOUS, 1985. Utchi LITAID species or group : An Album of Guangdong Litchi Varieties in chinensis CIRAD-FLHOR s/species or s/group : Full Colour. Guangdong, Chine, Guangdong form or cultivar : Kwai Mi () Provincial Agricultural Commission, Guang­ dong Provincial Scientific and Technical Commission, 78 p. BATIEN D., 1984. COLLECTION: P92 number code 1 1 1 11 Lychee varieties, Agfacts H 6.2. 7. Dep. Agri. N.S.W. 15 p.

General Flower Characters BERTIN L., 1992. Le litchi a la Reunion : possibilites d'amelio­ 1 Date of onset of flowering : Normal (Ref.: Kwa·i Mi) ration des exportations. Memoire de fin 2 Date of bud burst : Normal ± 10 days d'etudes. ISTOM - CIRAD/FLHOR Reunion, 3 Calyx colour : Green 52 p. + annexes. 4 Hairiness : Downy COMMON J.H., 1993. 5 shape : Pointed Contribution a I' identification varietale chez 6 Sepal suture : United le litchi (Litchi chinensis Sonn.). Analyse des 7 Nectariferous disk colour : Orange caracteres morphotaxonomiques floraux. 8 Anther shape : Wide 0.7 < L/L Memoire de fin d'etudes ISTOM 9 Number of : 6 or 7 CIRAD/FLHOR Reunion, 41 p. + annexes. 10 Peduncle length (mm) : Average 0.30 < h s; 0.65 HILU K.W., 1989. Male Flowers Taxonomy of cultivated plants. IBPGR Training Courses: Lecture Series 2, Scientific 11 Calyx size (mm) male fl : Average 3.0 < d s; 3.5 Management of Germplasm: Characteriza­ 12 General shape male fl : No answer tion, Evaluation and Enhancement. H.T. 13 Boll shape male fl : Bell Stalker et C. Chapman, IBPGR, Rome, Italy, 14 Boll constriction male fl : No p. 33-40. 15 Position of nect. disk male fl : Not enclosed 0 < h LIANG J., 1981. 16 Disk protuberance male fl : Protuberant 0.25 < h s; 1.00 Le litchi, origine, utilisation et developpement 17 Filament length (mm) male fl : Average 3 < L s; 4 de sa culture. Jour. d'Agric. Trad. et de Sota. 18 Anther length male fl : Average 1.10 < L s; 1.25 Appl. XXVIII (3-4) : 259-269. 19 Peduncle colour male fl : Pink MENZEL C.M., SIMPSON D.R., 1986. 20 Ovary length male fl : Long 0.4 < L Lychee cultivars - Description and perfor­ 21 Ovary width male fl : Wide 0.9 < mance of major lychee cultivars in subtropi­ 22 Stylus length (mm) male fl : Very long 1.20 < L cal Queensland. Queensland Agricultural 23 Stylus colour male fl : Greenish Journal : 125-136. Female Flowers NORMAND F., 1990a. Compte-rendu de la prospection locale de lit­ 24 Calyx size (mm) fem fl : Large 3.5 < L s; 4 chis 1988. LI-RE n· 38, IRFA-Reunion, 8 p., 25 General shape fem fl : Bowl internal document, ARC 20-002. 26 Boll shape fem fl : Bell 27 Boll constriction few fl : No NORMAND F., 1990b. Deux ans d'observation de la collection de 28 Position of nect. disk fem fl : Not enclosed 0 < h litchis Reunion BPL07. LI-RE n" 39, IRFA - 29 Disk protuberance fem fl : Barely protuberant h s; 0.25 Reunion, 22 p., internal document, ARC 20- 30 Filament length (mm) fem fl : Short L s; 1.0 001. 31 Anther length fem fl : Short L s; 1.15 32 Peduncle colour fem fl : Greenish NORMAND F., BOUFFIN J., PERRIER X., HUAT J., 33 Ovary length fem fl : Short L s; 0.7 1990. 34 Ovarywidth fem fl : Narrow I s; 1.2 La reconnaissance varietale chez le litchi (Litchi chinensis 35 Stylus length (mm) fem fl : Short L s; 1.8 Sonn.) a l'aide de carac­ 36 Stylus colour fe m fl : Greenish teres floraux, IRFA Reunion, 10 p., internal document, ARC 20-265. 37 Stigma length (mm) fe m fl : Average 1.2 < L s; 1.8 38 Stigma shape fem fl : Bent PERRIER X., TEZENAS du MONTCEL H., 1988. 39 Stigma shape fem fl : Greenish MUSAID: a computerized determination sys­ tem. In : Proceedings of INIBAP workshop, Los Banos, Philippines, 5-7 September Figure 1 1988. Ed. Jarret, 218 p. Description sheet for cv Kwai'. SINGH L.B., SINGH U.P., 1954. Mi floral characters produced The litchi. Saharanpur, lnde, Lucknow, U.P., by the LITA/0 softwareprogr am. 87 p. • •••

382 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards Inventory of Tropical Fruit Trees in. Ce.ntral America and the West 1.ndies

G. BARBEAU • • • .. t!I • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Projet fru itier regional MAE/IICA There is a broad range of tropical fruit genetic resources in Central Trinidad and Tobago America and the non-French West Indies. Some relatively isolated species could be developed. ° Fruits, vol. 49, n 5-6 •••••••••••••••••@•••••••••••••o•••••••••••••• p. 383-389 (English) p. 469-4 7 4 (French) introduction trated the West Indies and Central America, where Mangifera indica is the Many different tropical fruit species were only species present. brought to the non-French West Indies There are now very many cultivars in from India, Indonesia, Central and South Central America, the Greater Antilles and America during the colonial period. especially the Dominican Republic that Although some of these species have only were obtained by seed propagation of a curiosity value, others are now of eco­ monoembryonic varieties. nomic importance. There is not so much diversity in the Central America was long the private Lesser Antilles, despite the variety of ground of Spain and other foreign powers types; cv Julie is dominant in many areas. had but temporary and geographically There is more diversity in the south. In limited effects in the region. Some fmits, th th such as citrus and mango, only recently the 17 and 13 centuries, English and infiltrated this area from the West Indies. Dutch colonialists introduced many The plethora of local fruit resources com­ Indian varieties (in Trinidad and Guyana) pensated for this unfortunate isolation. and Indonesian varieties (in Guyana and Surinam). Mango varieties adapted to a Some Caribbean countries (Guyana, broad range of ecological conditions, Surinam, Trinidad), because of their geo­ from zones with a marked dry season in graphical positions, also have genetic the southwestern coastal region of resources native to the Amazon region. A Trinidad to the wet tropical conditions of number of locally-appreciated species Surinam, can now be found. This have yet to be investigated in detail. resource potential would be interesting to Knowledge of fruit genetic resources is tap, particularly to obtain varieties resist­ essential for the development of agricul­ ant to anthracnose and fruitflies. In tural diversification programs. Various fac­ Trinidad, this promising potential is high­ tors such as deforestation and financial lighted by the fact that mangoes are rarely shortages are threatening the survival of infested even though 16 different Ana­ some species. A germplasm inventory of strepha fruitfly species are present and fruit trees from Central America and the ready to infest a variety fruit species. non-French West Indies is now available; A limited list of mango varieties found in this is an important step towards the con­ Trinidad, Guyana and Surinam is given in servation and utilization of this plant Table 1. material. cashew Anacardiaceae Cashew (Anacardium occidentale) ong1- nates from tropical regions of the mango Americas. It is well adapted to Central Mangos were introduced in Brazil in the American and West Indian areas with a 17th century and then they slowly infil- marked dry season. They are cropped in

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 383 • e BARBEAU

A close species, the red mombin Table 1 (S. mombin L. = S. lutea), is a tall slender Varieties of mango (Mangifera indica) found in Trinidad, Guyana tree that produces clusters of yellow fruit and Surinam. during the rainy season. In Central America, the fruit is cooked before con­ Trinidad Guyana Surinam sumption. It grows better in wet tropical Bombay Aromanis areas and is found in natural forest stands Axe Buxton spice Buxton spice in Trinidad, Guyana and French Guiana. Buxton spice Colonial Bank Cayenne In Surinam and Guyana, the fruit is har­ Calabash Dor Goleth vested and used to produce fresh juices Ceylon Graham Julie that are sold in restaurants and by street Cheese Hard spice Pandiet vendors. Coden Julie Roodborstje Doux deuce Number 11 Tete, etc. The ambarella (S. cytherea Sonner), from Graham Peter Polynesia, is another species of this Julie Primrose Long Sand spice that should be mentioned. It is not very Peach Vauraj, etc. common in Central America, but is wide­ Sandersha spread in the Antilles, and some small Starch countries such as Grenada and Saint Vert Vincent now export ambarellas (mainly to Zabricot, etc. Trinidad and Canada). Small volumes are also exported to UK and the Netherlands. commercial plantations in Salvador and A varietal breeding programme is under Panama and in a small orchard in way in Grenada. There is an interesting Trinidad. Improved varieties from Brazil and promising dwarf type from Southeast were introduced in Nicaragua in the Asia that has been introduced in Trinidad. 1970s, and in Saint Lucia more recently. They are currently being bred, especially for anthracnose resistance, in Panama and Annonaceae Trinidad. In a germplasm repository in This is well represented in Central Salvador, there is a type that produces a America and the West Indies and many black false fruit! The species is commonly important species originate from these found growing wild on plateaus and regions. Soursop (Annona muricata L.) white sandy savannas of Guyana, and and custard apple (A. squamosa L.) are cashew nuts represent a considerable common garden species. There are com­ resource for local Indians. mercial soursop orchards in Costa Rica, Panama and Surinam. It is abundant in Spondias species Grenada and Saint Vincent where it is grown in gardens or intercropped; The Spanish plum (Spondias purpurea L.) Trinidad is the main export market. is ve1y common on the Pacific coast of Varietal improvement of soursop is under Central America and in regions of the way in a few countries, including Costa Greater Antilles with a very marked dry Rica. season, its preferred environment. It is a wide-crowned upright tree. Other less common Annonaceae species are garden crops or grow subspontane­ About a dozen different types are known ously, including the Guatemalan annona and have been propagated by smallhold­ or papauce (A. divesifolia Safford) and ers. It is easily propagated from large bullock's heart (A. reticulata Mill.). The branch cuttings at the end of the dry sea­ former species produces large delicious son, thus explaining why they are used pinkish-green fruit, which unfortunately for fence posts by farmers. Red plums are tends to crack during ripening, making it abundant during the dry season, but susceptible to ant infestation. The latter unfortunately this fruit is often infested species produces smooth yellowish-red with fruitfly larvae (Anastrepha spp.); fruit and is hardier and very drought these pests, however, are not present in resistant; it is sometimes used as root­ Grenada or Saint Vincent. stock for soursop.

384 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES •

Soncoya (A. puipurea Moc. and Sesse) is Other Garcinia species have been found much less common. This self-sown spe­ growing wild, and in botanical gardens cies is found in natural forest stands of and private collections, especially in Central America and Trinidad. It has ve1y Salvador, Nicaragua and Trinidad. They large leaves and orange-fleshed fruit cov­ are difficult to identify. ered with pyramid-shaped protuberances. As A. reticulata, it has graft-compatibility Lauraceae with soursop and can therefore be used as rootstock for this species. Avocado (Persea americana Mill.) origi­ nates from Central America and the West ote that pond apples (A. glahra L.) are Indies. There is a wide range of types in often found in wet habitats and some­ these countries since each tree propa­ times used as rootstock for corossol in gated from a pit is actually unique. The poorly drained areas. species has a high degree of plasticity In addition, there are other yet uninvesti­ considering the fact that there are three gated species of this genus in Central races, with many intermedia,y hybrids, America and the West Indies (A. montana adapted to various ecological conditions. Macfad., A. scleroderma Saff. , A. paludosa It is quite possible to select economically­ Abul., etc.). interesting new cultivars from wild forms; Another genus, Rollinia, is also present many countries are thus propagating their and grown for its fruit. Biriba (R. deliciosa own improved varieties for domestic and Saff. = R. pulchrinervia DC.) is a garden sometimes export markets. A few exam­ crop in Trinidad, with several endemic ples are given in Table 2. species in Guyana and French Guiana, Avocado harvests could be extended including R. mucosa Baill. and R. pulchri­ throughout the year by promoting some neruia A. DC .. of these locally-improved varieties, pro­ Unfortunately, many annona fruit crops ducers would thus be able to meet the are plagued by weevils (Cerconata ano­ domestic market demand. nella, Bephrata maculicollis) , whose lar­ In Central America, many so-called wild vae attack the seeds and cause fruit rot. types have been grouped together under This is not yet a problem in Grenada and the umbrella of Persea schiedeana ess .. Saint Vincent. Californian research scientists have con­ In response to a request by Andean coun­ ducted many studies on these popula­ tries, the International Plant Genetic tions, seeking forms that are resistant to Resources Institute (IPGRI), which is collar rot (Phytopthora spp.). based in Palmira (Colombia), intends to begin a project on Annonaceae species. Lecythidaceae Brazil nuts (Bertholletia excelsa Hum. and Guttiferae Bon.) and other similar nuts (Lecythis The mangosteen C Garcinia mangostana L.) is relatively unknown in Central America and the West Indies. To our Table 2 (Persea americana knowledge, there has only been one Varieties of avocado Mill.) obtained in local breeding programmes. mangosteen orchard (about 4 ha; 50 years old) in these regions; this was at Cukra Nicaragua Guyana Dominican Republic Hill, on the Atlantic coast of Nicaragua, which might have been destroyed by Apante Chanasue Gripina 12 cyclone Johan in 1988. This species is Campos azules Hercules Gripina 5 also present at (Honduras). We Chalet Hosororo 5 Melendez 2 Laura Hosororo 6 Semi! 31 were quite surprized to come across sev­ Mangei'io Khan Semi! 34 eral stands of old mangosteens in Masatepe Mahabir Trinidad; they are flourishing and seem to Moyei'io Matthew's Ridge be quite productive. The Trinidad Rosario Pierre Ministiy of Agriculture has set up nurser­ Silva ies to propagate this species. Ticomo

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 385 • e BARBEAU

zabucajo Aubl., Lecythis elliptica H.B.K., wet tropical regions. There are white- and etc.) from the Amazon region are not very yellow-fleshed varieties but, due to the common in Central America and the West fact that it is only propagated vegeta­ Indies, and little is known about them. tively, there is little variability concerning Nevertheless, several stands of L. zabu­ the other characters. cajo were introduced in Trinidad about There have been several studies on this 50 years ago and they are still in excellent species, notably at UWI Trinidad, in shape. The nuts are consumed locally and Jamaica and Barbados. Because of its sometimes sold in supermarkets. This high market value, West Indian countries species would be of interest for rehabili­ are interested in introducing other varie­ tating poorly drained lands in wet tropical ties obtained from their origin and diver­ areas. sity centres. Seeded breadfruit (Artocarpus altilis var. Malpighiaceae semin[fera) is not quite as interesting, The Barbados cherry or West Indian only the grilled or boiled seeds are con­ cherry (Malpighia glabra L.) originates sumed. form the Antilles and is cropped in Barbados, Guyana and Surinam; it is also jackfruit and chempedak quite common in the rest of the West Jackfruit (Artocarpus heterophyllus Lam.), Indies. in contrast to breadfruit, has whole or This cherry is not as widespread in only partially denticulate leaves. Single Central America. Improved varieties have fruits can weigh as much as 20 kg, and been bred in Puerto Rico, the Dominican they develop on the trunk and largest Republic and Surinam. Many other spe­ branches (similar to cacao). This species cies have been described in the Antilles is only a curiosity in Central America, but and Central America, i.e. M. angustifolia is grown throughout the West Indies L., M. aquifolia L. and M. coccife ra L. in where the fruit is mainly consumed by the Lesser Antilles, M. cnide Spreng. in Hindus (in Guyana and Trinidad). the Dominican Republic, M. .fucata Ker Chempedak (Artocarpus integer [Thun.) Gaw!. in Jamaica, M. incana in Honduras, Mer.) is smaller than the jackfruit tree. etc. This highlights the considerable The fruit peel and pulp are yellow; the breeding potential of this fruit. fruit has a banana-like smell and taste.

Moraceae Myrtaceae All fruit tree species in this family that are This large family can be found in the of interest in Central America and the tropics and subtropics, adapted to almost West Indies originate from Southeast Asia all biotopes. Many cropped species origi­ and Polynesia. nate from tropical regions of the Breadfruit is the first species of this family Americas, and others are from Southeast to be introduced in the West Indies; it Asia. was brought by Captain Bligh on the guava famous Bounty. Saint Vincent celebrated the 200th anniversary of this visit in Psidium guajava L. is a species with high November 1993. plasticity, producing relatively large fruit of various shapes, with a yellow to red seeded and seedless breadfruit pulp. It is cropped in commercial Breadfruit (Artocmpus altilis [Park.] orchards in some countries, to be pro­ Fosberg) is grown everywhere in the cessed on a small industrial scale. It was West Indies where it is consumed tradi­ formerly the focus of a substantial tionally. It is also exported by Saint Lucia improvement programme in Trinidad (cv and Jamaica. Breadfruit is less common in Centeno Prolific), and more recently in Central America where it is only found in Cuba, where it is very popular and crop-

386 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES e

ped in about 2 OOO ha of orchards. Two Java plum high-yielding dwarf varieties are now available to producers. Guava is very sus­ The Java plum or Black jamoon ceptible to fruitflies, which limits its pres­ (Syzyg ium cumini Skeels) is called jam­ ence on fresh fruit markets. bolan in the French West Indies; it origi­ nates from Southeast Asia. It is common Costa Rican guava in the English West Indies where it serves as a windbreak for other crops. The dark­ Ps idium friedrichsthalianum is native to purple cherry-sized fruit, which grows in Central America. The fruit is more sour clusters, is used to make jam and wine. than that produced by P guajava; it is Some types produce more oblong-shaped greenish-brown with white aromatic pulp. fruit. This sour Costa Rican guava is mainly found in Central America where it is grown in gardens and processed into Oxalidaceae fresh juice. The shape and size of the fruit varies. It should be investigated in terms carambola of its potential development, in the light of its ve,y special flavour (quite different Carambola (Averrhoa carambola L.) origi­ from Psidium guajava) and low suscepti­ nates from Indonesia. The tree is average bility to fruitfly infestation. sized, very ramified, and produces yello­ wish-orange fruit with five clearly-marked malay apple sides. Its popularity is increasing on world markets. Syzyg ium malaccensis Merr. and Peny is limited to the Atlantic side of Central Although this species was introduced in America where the moisture conditions many regions, it is now only found in are suitable for its development. This spe­ large numbers in some countries, i.e. cies, which originates from Malaysia, is coastal areas of Guyana (Pomeroon), much more common in the West Indies Surinam, and southern Nicaragua (Rivas, where it is grown in gardens, and some­ Buenos Aires, Potosi). Carambola is prop­ times serves as a windbreak for other agated sexually and there is considerable crops (Trinidad). In Grenada, it is called variability in some characters, especially the «French cashew» because of the simi­ fruit size, shape and size of the sides, larity between its fruit and the false fruit peel and pulp colour and the fruit oxalic of cashew trees. The fruit is mainly con­ acid content. The latter character differen­ sumed fresh, but sometimes as juice. tiates sweet and sour varieties. The wide There is very little genetic variability. The array of intermediate varieties are now seeds are polyemb1yonic and the mother­ considered for varietal improvement, and plant characters are reproduced in the vegetative propagation techniques are progeny. Two types are propagated com­ being developed in Trinidad. mercially from cuttings in Trinidad, the In Nicaragua, the fruit is mainly processed fruit of one is red and the other is strea­ into juice. It is used to make a prune sub­ ked with white. The fruit is often infested stitute in Guyana, and a date substitute in by fruitflies. Surinam. jamelac Jamelac (Syzygium samarangense Merr. bilimbi or pickle and Perry) is found in botanical gardens Bilimbi (Averrhoa bilimbi L.) is a small and private collections as a curiosity; it is upright tree originating from Malaysia. commonly grown in gardens in Surinam. The green acidic fruits resemble pickles Jamelac was probably brought from and hang from the trunk or the main Southeast Asia by the Dutch, or by branches in tight clusters. It is found Javanese people that now live in Surinam. everywhere in the West Indies and The fruit is white or pale-pink, with a Central America but its distribution is lim­ Chinese-hat shape, and is used to make ited. The fruit is used in pickled form and delicious candied fruit. for the preparation of Hindu dishes

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 387 •• • BARBEAU

(Trinidad, Guyana). Small plantations shady conditions; a lemon hybrid, a blend have recently appeared in Trinidad and of West Indian lime (C. aurantifolia the crops are mainly processed. Swing.) and Tahiti lime (C. latifolia Tan.), was also bred for resistance to anthrac­ nose, which was devastating lime crops. Rhamnaceae There are also several types of cv jujube Cleopatra mandarins, which could replace cv Sour Orange as rootstock, since this Most jujube species ongmate from Asia latter variety has been affected by tristeza (Ziziphus jujuba Mill., Z. mauritania virus and can no longer be used. Lam.) or Africa (Z. lotus Lam., Z. spina­ christi Willd.). Only Z. mauritania is well acclimatized to the tropical conditions in Sapindaceae the Americas, and especially the West Indies. It often grows as a shrub, some­ rambutan times a large tree, and is found in dry and Rambutan (Nephelium lappaceum L.), moist zones. Small-fruit varieties are most originating from Southeast Asia, was common, but improved large-fruit varie­ introduced in many parts of Central ties, propagated by budding, are also America and the West Indies. It is present found. The fruit is consumed in fresh or in some repositories, i.e. in Honduras and processed form (sold in supermarkets in Costa Rica, but rarely cropped; neverthe­ Trinidad). It could be developed for crop­ less, there are a few small orchards in

ping in dry areas of the West Indies. Costa Rica and Trinidad, and experiments are under way in Honduras. Fruit from some scattered trees are marketed in Rubiaceae Guyana and Surinam. This species is Genipa ( L.) is found mainly propagated from seed and shows in natural forest stands in wet tropical very high diversity for characters such as regions of the Americas and the West yield, fruit size, colour and pulp adhesion Indies. It also grows in dry areas, but this to the seed, etc. A rambutan breeding programme is under way in a small is not very common. The rounded, grey­ ish, hard-sl1elled fruit grows to various orchard in Trinidad. sizes and its quality also varies markedly. Large-fruit varieties with bulky sweet pulp pulasan have been bred in the Dominican Pulasan (Nephelium mutabile Blume), Republic. also originating from Southeast Asia, is even less common than rambutan. To our knowledge, only a few trees are grown Rutaceae on a property in Trinidad. Citrus fruits were first introduced in the The latter two species seem to perform th West Indies at the end of the 15 century. suitably in wet tropical areas of the Citrus research has never been as impor­ Americas and could be cropped on a lar­ tant in Central America and the West ger scale. Indies as it has been, and continues to be, in Florida and some other states in akee USA. Nevertheless, the West Indies have Akee (Blighia sapida Koenig), native to also been instrumental in promoting Western Africa, has been introduced in citrus fruit through the development of many areas, but is still only a curiosity in grapefruit in the the 181h century and cvs the West Indies. In Jamaica, on the other Ortanique and Ugli in Jamaica in the zorh hand, akees are very common garden century. The Imperial Agricultural College trees and the fruit is used in local dishes. of Trinidad used to conduct investigations on citrus fruit. There is still some evi­ honey-berry dence of this research in gardens and Honey-berry (Meliccoca bijuga L.) is cacao plantations: a sweet orange variety, endemic to northern regions of South Cocoa, is known for its adaptation to America. It is a large hardy tree grown in

388 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / GENETIC RESOURCES •

gardens, hedges between properties, or canistel or egg-fruit intercropped. The fruit is green, grows in clusters and there is considerable variabil­ Canistel (Pouteria campechiana Baehni = ity in its shape, pulp sweetness and bulk, Lucuma nervosa D.C. and L. salicifolia which can be explained by the fact that it H.B.K.) has golden-coloured fruit with is dioecious. The fruit is sold everywhere abundant sweet fruit whose texture in markets and along roadsides. It is gen­ resembles cooked egg yolk. There is mar­ erally consumed fresh, but sometimes as ked variability in tree size, growth habit a cool drink in Central America. and fruit size. It has not been the focus of any breeding programmes. It is almost always propagated from seed. Sapotaceae balata This family is ve1y well represented in Balata (Manilkara bidentata Dub.) is a Central America and the West Indies. forest species in the Antilles. It is propa­ Some species are sold widely on domes­ gated in nurseries in Trinidad and planted tic markets, but there is veiy little export in gardens. The milky-pulped fruit is con­ trade. One exception concerns sapote sumed fresh and sold in markets and pulp, which is processed at a plant in along roadsides. Guatemala and exported to USA. A few of the most important species are discussed other species found in Central below. America and the West Indies mammee sapote There are some other noteworthy native Mammee sapote (Calocarpum sapota Central American and West Indian species Merr. C = C. mammosum Pierre) is abun­ that are not propagated but their fruit is dant throughout Central America, Cuba harvested: and the Dominican Republic. There are - green sapote (Calocarpum viride Pitt.) many different types with high variability in Central America, for some characters, such as fruit size and - several Ch rysophyllum species in the shape, pulp colour and quality. There has Antilles, Guyana and French Guiana, been veiy little research focusing on this - several Lucuma species in Cuba and species, and it is almost always propa­ Puerto Rico, gated from seed. Sapote crops are suscep­ - several Pouteria species in Trinidad, tible to fruitfly infestation. Guyana and French Guiana, including the penny piece (Pouteria multiflora) in sapodilla Trinidad, and P. macrophylla and P. guia­ Sapodilla (Manilkara zapota Van Royen = nensis in Surinam. Achras sapota [Mill.] Fosb.) has the same Sweetberiy (Synsepalum dulcificum characteristics as the mammee sapote. Its Daniell), which is found in botanical gar­ distribution also includes the Lesser dens and private collections, is also note­ Antilles and Trinidad, where some varie­ worthy. In the 1930s and 1940s, many ties have been identified (De Meillac, studies were carried out in USA on this Abdool) and propagated by budding. It is species with the aim of developing a susceptible to fruitfly infestation. sugar substitute. star apple In response to a request by Central American countries, the International Star apples (Ch sophyllum cainito L.) are ry Plant Genetic Resources Institute (IPGRI), found throughout Central America and which is based in Palmira (Colombia), the West Indies. There are two main intends to begin a project on Sapotaceae groups, one with purple fruit and pulp, species of Central America. • the other with green fruit and white pulp. It is propagated from seed. Some types differ with respect to fruit size and shape. A large-fruit variety was bred in Cuba. It is also susceptible to fruitfly infestation.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 389 Nuclear Genome Size Variations in Citrus

P. 0LLITRAULT, D. DAMBIER •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR, BP 5035 Systematic analysis of genome sizes in Citrus is important 34032 Montpellier cedex 01 fo r programmes involving gene mapping and plant breeding France F. LURO via sexual crossing. SRA San Giuliano ••••••••••••••••••••••••••••••••••••••••••• ••• CIRAD-FLHOR/INRA 20230 San Nicolao Corsica introduction thousand nuclei per sample were then analysed by Fascan flow cytometry con­ C. DUPERRAY Numerical taxonomic studies (BARRETT INSERM, U291 nected to the Lysis 2 software pro­ and RHODES, 1976) and molecular marker gramme. The size of the cv Tahiti lime 99, rue Puech Villa et al., 34090 Montpellier analyses (GREEN 1986; OLLITRA LT nuclear genome was evaluated relative to France and FAURE, 1992; YAMAMOTO et al., 1993) chicken e1ythrocytes (2.33 pg/2C) so as to have demonstrated that the wide variety determine absolute sizes of genomes for of cultivated Citrus species can be clearly each diploid cultivar considered. The data Fruits, vol. 49, n°5-6 grouped around three basic taxa since were analysed using hierarchical analysis p. 390-393 (English) Citrus medica (citrons), Citrus reticulata p. 475-476 (French) of variance to determine possible size (mandarins) and Citrus grandis (pum­ variations within each species and assess melo) are generally considered to be the interspecific diversity. three ancestral species of currently culti­ vated Citrus. Despite the number of sym­ patric zones, this level of structuring sug­ results gests that many factors limit interspecific The absolute size of the cv Tahiti lime recombination. Some of these limitations nuclear genome was estimated at are genomic, as shown by the structural 1.17 pg/2C for 2N=3X=27 chromosomes. heterozygosity (translocations, inversions, The coefficients of variation (c.v.) con­ etc.) that has been noted in interspecific cerning the GO-G 1 peaks for the different T et hybrids (RAGHUVANSHI, 1969; GMIIER diploid samples was around 2.5-3% for al., 1992). In the present study, we loo­ most of the cultivars (Fig. 1). However ked for a quantitative component to the c.v. of peaks were systematically account for the structural differentiation w higher for some varieties of sweet orange bet een nuclear genomes of various sweet (C. sinensis), lime (C. aurantifolia) Citrus species. and mandarin, which resulted in higher variance for the genome size estimates. material and methods In six cases, there were significant varia­ tions in genome sizes (not higher than The relative nuclear genome sizes of four 3%) between cultivars of the same species to five diploid cultivars of the eight main (Table 1) cultivated Citrus species were assessed (Table 1). For each cultivar, evaluations There was also ve1y marked interspecific relative to the control triploid cultivar cv variability, i.e. reaching 10% between Tahiti lime were done in triplicate. Leaf mandarin and citron (Figs. 1 and 2), the samples from the test cultivar and control smallest and largest Citrus genomes were minced finely in PBS buffer supple­ respectively. The other species were divi­ mented with dithiothreitol (1 mg/ml), ded between two other intermediate-sized Triton XlOO (0.3%) and RNAase groups. One included sweet orange and (lo-3 U/ml). After filtration, 0.3 ml of the sour orange (C. aurantium) and the other nuclear suspension was mixed with lemon (C. lemon), lime, pummelo and 0.3 ml extraction buffer supplemented grapefruit (C. paradisi). This division into with propidium iodide (200 mg/ml). Two four separate groups, revealed by the

390 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY e

Table 1 Estimated relative and absolute sizes of nuclear genomes in Citruscu ltivars (/cv Tahiti lime). Mean values for each species are presented in bold. H.G. = homogeneous groups determined by the Newman-Keuls test. F and p = statistics associated with the Newman-Keuls test.

Cultivars Relative Absolute H.G. Intra specific size size variability

C aurantium (Sour orange) 0.643 0.752 Bigaradier Goutou 0.642 0.751 a F = 2.39 Bigaradier Granito 0.644 0.753 a p = 0.14 Bigaradier Maroc 0.641 0.756 a N.S. Bouquetier Nice 0.646 0.756 a

C medica (Citron) 0.696 0.814 Corse 0.696 0.815 a F = 10.71 Digite 0.697 0.816 a p < 0.01 Etrog 0.702 0.821 a Poncire commun 0.690 0.807 b

C Umon (Lemon) 0.665 0.778 Lisbonne 0.672 0.786 a F=ll.93 Sweet 0.665 0.778 b p < 0.01 Eureka 0.664 0.777 b Meyer 0.660 0.772 b

C aurantifo lia (Ume) 0.659 0.771 Brazil Sweet 0.646 0.756 a F = 9.08 Kirk 0.666 0.779 b p < 0.01 Mexican 0.666 0.779 b .. Rangpur 0.660 0.772 b

C sinensis (Sweet Orange) 0.647 0.757 Parson Brown 0.646 0.756 a F = 0.02 Shamouti 0.646 0.756 a p = 0.99 Washington Navel 0.647 0.757 a N.S. Valencia late 0.647 0.757 a

C paradisi (Grapefruit) o.666 0.779 Star Ruby 0.660 0.772 a F = 7.36 Cecily 0.665 0.778 b p = 0.01 Marsh 0.669 0.783 b Thomson 0.670 0.784 "b

C grandis (Pum.melo) 0.669 0.783 Kao Pane 0.656 0.768 a F = 25.01 Pink 0.666 0.779 b p < 0.01 Seedless 0.673 0.787 be .. Inde 0.673 0.787 be Sunshine 0.679 0.795 C

C reticulata (Mandarin) 0.630 0.737 Satsuma Wase 0.623 0.729 a F = 9.14 Willow-leaf 0.624 0.730 a p < 0.01 Cleopatra 0.627 0.733 a Ponkan 0636 0.745 b

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 391 •• • OLUTRAULT et al.

1 2

GO/Gl control 60 GO/Gl Corse citron G2/M Corse citron Etrog citron

Willow- leaf mandarin GO/Gl 'ii) ·v GO/Gl control u u Pink ::, ::, G2/M z z pummelo <..., <..., Pink pummelo 0 0 .... (l) ..0 ..0 § ::, z GO/G l z control GO/Gl Satsuma G2/M mandarin Satsuma mandarin 0 0 200 400 600 800 1000 Relative Quantity of DNA Relative Quantity of DNA

Figure 1 Estimated relative sizes of nuclear genomes in three Newman-Keuls interspecific test results, lata). Note that the pomelo variety Star diploid cultivars relative to was confirmed by the quadrimodal distri­ Ruby, with a smaller nuclear genome cv Tahiti lime. butions of single observations (Fig. 3). than the other varieties of this species, was developed in a gamma irradiation Figure 2 programme (HENSZ, 1960). Genome size differentiation discussion in cv willow-leaf mandarin Interspecific diversity was relatively high This is the first systematic genome-size and cv Etrog citron. and the genome size variations noted are study that has been carried out in Citrus. in line with phylogenic hypotheses. The In ARUMUGANATHAN and EARLE c. sinensis, extreme values obtained for 0991) estimated a genome size of C. medica and confirm their ancestral 0.76 pg/2C in comparision to chicken C. reticulata roles in cultivated the estimated e1ythrocytes, but there are no other previ­ Citrus; genome sizes for secondary species is in ous results for the other species covered agreement with their assumed hybrid ori­ here. Diploid cultivars seemed to have gins (OLLJTRAULT and FAURE, 1992). The relatively small genomes (0.73-0.82 pg/ quantitative differentiation in the three 2C). ARUMUGANATHAN and EARLE (1991) basic taxa is evidence of advanced evolu­ obtained similar results for many other tion towards real speciation. It undoubt­ fruit trees, e.g. apricot (0.61 pg/2C), edly helped maintain high linkage dis­ peach (0.54 pg/2C), mango (0.91 pg/2C) equilibrium in Citrus and could explain and pear (103 pg/2C). some non-Mendelian segregations that In Citrus, significant intraspecific diversity have been observed in the progeny of was noted, in species that have evolved interspecific hybrids (OLLITRAULT and solely by mutations (C. lemon and C. par­ FAURE, 1992). This differentiation and its adisi) and those that have diversified by impact should be taken into account in sexual crosses and mutations (C. grandis, gene mapping studies and plant breeding C. auran[!fo lia, C. medica and C. reticu- programmes involving sexual crossing. e

392 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY •

Number of observations 10

5

0 Genome sizes

C. medica II C. grandis a • ••• C. paradisi a •••• C. lemon D ODD D C. aurantifolia C. sinensis a :=::======-=-=====;.:;===:;;;; C. aurantittm a ---- · -===;;;;.-;;=-;;:;;-; C. reticulata § ElEl

0.72 0.74 0 76 0.78 0.80 0.82 Figure 3 Variations in the size Homogeneous groups Mean nuclear genome sizes (Newman-Keuls test) of nuclear genomes in the Citrus genus (pg/2C).

HENSZ R.A., 1960. references Effect of X-ray and thermal neutrons on citrus propagating material. J. Rio Grande Val. Hort. ARUMUGANATHAN K., EARLE E.D., 1991. Soc. 14: 21-25. Nuclear DNA content of some important plant species. Plant Mol. Biol. Rep. 9 (3): 208- OLLITRAULT P., FAURE X., 1992. 218. Systeme de reproduction et organisation de la diversite genetique dans le genre Citrus. BARRET H.C., RHODES A.M., 1976. Proceedings of the international conference A numerical taxonomic study of affinity rela­ "Complexe d'espece, flux de genes et res­ tionships in cultivated Citrus and its close sources genetiques", Paris, France, 8-10 relatives. Syst. Bot. 1: 105-136. January 1992. Paris, France, Bureau des Ressources Genetiques ed., p. 133-151. GREEN R.M., VARDI A., GALUN E., 1986. The plastone of Citrus. Physical map, varia­ RAGHUVANSHI S.S., 1969. tion among Citrus cultivars and species, and Cytological evidence bearing on evolution in comparaison with related genera. Theor. Citrus. In: Proc. 1st Citrus Symp., Chapman Appl. Genet. 72: 761-769. ed., Univ. of California, Riverside 1: 207-214.

GMITIER F.G., DENG X.X., HEARN C.J., 1992. YAMAMOTO M., KOBAYASHI S., NAKAMURA Y., Cytogenetic mecanism underlying reduced YAMADA Y., 1993. fertility and seedlessness in Citrus. In: Proc. Phylogenic relationships of citrus revealed by of VII lnt. Citrus Cong., lnt. Soc. of RFLP analysis of mitochondrial and chloro­ Citriculture. p. 113-116. plast DNA. Japan. J. Breed., 43: 355-365.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 393 Optimized Management of Citrus Embryogenic Calli for Breeding Programmes

P. 0LLITRAULT, D. DAMBIER, •••••••••••••••••••••••••••••••••••••••••••••• C. CABASSON Embryogenic ea/Ii were obtained fo r 15 Citrus cultivars. CIRAD-FLHOR, BP 5035 Cryoconservation is now routinely used fo r long-term callus conservation. 34032 Montpellier cedex 01 France lsoenzyme and flowcitome try techniques are applied for callus

V. ALLENT caracterization. SRA, INRA/CIRAD-FLHOR 20230 San Nicolao •••••••••••••••••••••••••••••••••••••••••••••• Corsica

F. ENGELMANN ORSTOM, BP 5045 34032 Montpellier cedex 01 introduction ced by ovule culture. Depending on the France genotype, the ovule can produce friable Somatic emb1yogenesis is now widely calli and emb1yos, compact chlorophyll used fo r protoplast fusion and gene trans­ calli, or just embryos. The results of iso­ fer in Citrus breeding programmes. These enzyme studies (OLLJTRAULT et al., 1992b) latter techniques provide means to over­ ° and histological analyses (CABASSON, 1993) Fruits, vol. 49, n 5-6 come some obstacles concerning genetic p. 394-397 (English) revealed that friable ernb1yogenic calli are structures of cultivars and considerably p. 477-478 (French) of nucellar somatic origin. Six months to expand the genetic base available to 1 year after induction, such calli can be breeders. However, they require con­ propagated on media without hormones. trolled management of embryogenic calli, Callus strains have thus been obtained for the essential component of this system 15 cultivars representing a wide-range of (Fig. 1). The main methodological diversity within the Citrus genus. advances jointly developed by CIRAD­ Emb1yogenic calli have also been formed FLHOR, INRA and ORSTOM are pre­ around the hypocotyl in micropropagatecl sented. somatic hybrids (Photo 1).

embryogenic callus induction cryoconservation Friable embryogenic calli from polyem­ of embryogenic calli bryonic varieties are conventionally indu- A simplified technique for callus cryocon­ servation in liquid nitrogen has been developed (ENGELMANN et al., 1994) to overcome callus induction problems with some genotypes and reduce somaclonal variations that could occur during succes­ sive subculturing. Calli are pretreated in a solution containing 0.15 M sucrose and 10% DMSO, placed in a small freezing module (Nalge Company) and then into a freezer set at - so·c. When the callus reaches - 40"C, the cryotubes are plunged in liquid nitrogen. This technique is now Photo 1 routinely used for long-term callus con­ Formation of an embryogenic callus around the hypocotyl servation and can also be used to con­ of a micropropagated somatic serve duplicate fusion and transformation hybrid of C. deliciosa and products during regeneration and evalua­ F. japonica. tion processes.

394 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY •

Callus Protoplasts Gene transfer irradation Somatic hybridization

Ovule Callus suspension culture subculture subculture

Cryopreservation

Galactose or low Galactose or water potential low cell density

Obtention of proembryos and embryos

Shoot or embryo grafting Acclimatization

Cultivar Rootstock selection selection and propagation Figure 1 Management of embryogenic ea/Ii. controlling embryo 20 regeneration 18 A agar 8itek 8 8acto agar 16 Nucellar callus emb,yogenesis can be H agar H (Sigma) obtained on MURASl·IIGE and TUCKER E agar E (Sigma) 14 medium (1969) without exogeneous hor­ M agar M (Sigma) mones by modifying the form of the sug­ -� 12 ars contained in the medium (OLLITRAULT, � 15 15 g/l 0 10 10 : 10 g/l 1992), or reducing the relative water con­ 0.. tent in the callus by increasing the g lling 7.5: 7.5 g/1 8 5 : 5 g/1 agent concentration (Fig. 2). With this ·a original latter technique, emb1yos can be 6 obtained for callus strains that do not respond to medium sugar modifications. 4 • E7.5 Emb1yo development varies according to 3 the callus strain and culture medium: cot­ M7.5 85 810 yledona1y emb1yos in cv Chios mandarin 0 • A5 ("agar effect", Photo 2) and cv Sunki 40 50 60 70 80 90 100 mandarin (galactose, Photo 3), and Relative water content (%) globular emb1yos in lemon (galactose, Photo 4). The observed suspensor-type Figure 2 structures (Photo 5) suggest that the Control of embryogenesis in willow-leaf mandarin ea/Ii by modifying the type emb1yos are of unicellular origin. of agar and its concentration, in relation with the relative water content in the callus.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 395 • • OLUTRAULT et al.

isoenzyme characterization of calli With the aim of obtaining very early detection of somatic fusion products, Photo 2 Cotyledonary embryos eight enzymatic systems (A.D.H., I.D.H., of cv Chios mandarin. M.D.H., Sk.D.H., P.G.I., P.G.M. G.O.T. Agar effect: MT medium and L.A.P.) were analysed by electropho­ (MURASHIGE & TUCKER, 1969) resis in nucellar calli and emb,yos. This + 50 g/1 sucrose was the same as the technique used with + 4 g/1 phytagel. leaf and bark samples (OLLJTRAULT et al., 1992a). In most cases, enough polymor­ phism is observed to enable detection of heterofusion products (Photo 6). For most of the systems, the callus and emb1yo profiles were identical to those obtained with leaves from adult trees. The profiles Photo 3 differed with M.D.H. and A.D.H., and Cotyledonary embryos their expression seemed to be associated of cv Sunki mandarin. with the embryo maturation level. Profiles Galactose medium: for fully developed cotyledonary embryos MT medium obtained from calli were identical to + 30 g/1 galactose those of seed emb,yos, while globular + 2 g/1 phytagel. embryo profiles were identical to those of calli (Photo 7).

flow cytometry analysis of callus cell cycles and ploidy levels Flow cytometry can be used to assess ploicly levels and relative proportions of cells in the Gl and G2 phases in organs and calli in the growth phase (OLLITRAULT Photo 4 and MICHAUX-FERRIERE, 1992). This latter Globular embryos of lemon (galactose medium). information is important for isolating pro­ toplasts to be used in somatic fusion and genetic transformation. Cytometry was used successfully to analyse protoplast nuclei obtained from embtyogenic calli and stained with propidium iodide. In a prelimina1y study (Fig. 3), the highest lev­ els of cells in the G2 phase (20%) were obtained 18 clays after subculture. This technique is also useful for controlling ploicly stability of embryogenic calli for­ med around the hypocotyl of somatic tet­ raploid hybrids (Figure 1). This stability is necessar for mass propagation, involving Photo 5 y Embryo of C. deliciosa attached somatic ernb1yogenesis, of new root­ to an embryogenic callus by stocks obtained in somatic hybridization a suspensor-type structure. programmes. e

396 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY e references CABASSON C., 1993. Regeneration de la mandarine commune (C. G1 M deliciosa Ten.) par embryogenese somatique en milieu liquide. Fusion somatique et essais Gl de transformation genetique. PhD thesis, uni­ 2c versite de Montpellier II (U.S.T.L.), France, 124 p. T ENGELMANN F., DAMBIER D., OLLITRAULT P., 1994. Cryopreservation of embryogenic cell suspen­ G2 sions and calluses of Citrus using a simpli­ 4c fied freezing process. Cryo-Letters 15: 53-58. s

MURASHIGE T. TUCKER DPH, 1969. Growth factor requirement of citrus tissue Cell cycle followed by constantly culture. In: Proceedings of the First dividing meristem cells. (Gl) presynthesis International Citrus Symposium, University of phase, (G2)(T) postsynthesis phase, California, March 16-26, 1968. Riverside, (M) mitosis, (S) DNA synthesis. USA, Chapman HD, vol. 3, p. 1155-1161. (2c, 4c) DNA content of diploid cell nuclei in the G1 and G2 phases, OLLITRAULT P., 1992. respectively. Research of seedless "willow leaf" mandarin by gamma irradiation of nucellar callus. In: G2 Proc. of VII lnt. Citrus Cong., Acireale, Italy, 3-8 March 1992. lnt. Soc. of Citriculture, p. 113-116.

OLLITRAULT P., FAURE X., NORMAND F. , 1992a. Citrus rootstock characterization with bark and leaf isozymes; application for distingui­ Relative DNA content shing nucellar from zygotic trees. In: Proc. of VII lnt. Citrus Cong., Acireale, Italy, 3-8 March 1992. lnt. Soc. of Citriculture, p. 338-341. Figure 3 Analysis of the cell cycle OLLITRAULT P., OLLITRAULT F., CABASSON C., 1992b. of C. deliciosa embryogenic Induction de cals embryogenes d'agrumes par ea/Ii (10 days aftersubcu lture). culture d'ovules : determination isoenzyma­ tique de l'origine tissulaire des embryons. Fruits numero special agrumes (47): 204-212.

OLLITRAULT P., MICHAUX-FERRIERE N., 1992. Application of flow cytometry for Citrus gene­ tics and breeding. In: Proc. of VII lnt. Citrus Cong., Acireale, Italy, 3-8 March 1992. lnt. Soc. of Citriculture, p. 193-198.

EV EN F C E" EN F C EV EN F

Photo 6 Enzymatic profiles for 12 callus strains: 1 Sunki mandarin 2 Cleopatra mandarin 3 Sour orange 4 Shamouti orange 5 Nave/ate orange 6 Russ-Navel orange Photo 7 7 Star Ruby grapefruit Comparison of MOH expression in ea/Ii (C), 8 Mexican lime in vitro embryos (EV), nucel/ar seed embryos 9 LAC lemon (EN), leaves (F), for fo ur genotypes: cv Chios 10 Willow-leaf mandarin (G+) mandarin, cv Sunki mandarin, cv Shamouti 11 Willow-leaf mandarin (G-) orange and cv Star Ruby grapefruit. PGI 12 Chios mandarin.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 397 Facultative Apoximis, Spontaneous Polyploidization and Inbreeding in Citrus volkameriana Seedlings

P. 0LLITRAULT •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR, BP 5035 Th e genetic status of 135 seedlings of Citrus volkameriana 34032 Montpellier cedex 01 and relationships with plant vigor were analysed. France C. JACQUEMOND •••••••••••••••••••••••••••••••••••••••••••••• SRA San Giuliano INRA/CIRAD-FLHOR 20230 San Nicolao Corsica introduction Isozyme polymorphism was used to dif­ Fruits, vol. 49, n°5-6 ferentiate zygotic and nucellar seedlings. p. 398-400 (English) Citrus rootstocks are usually propagated Five heterozygous loci were analysed in p. 479-480 (French) by planting polyembryonic seeds that C. volkameriana using four enzymatic contain one zygotic embryo and several systems (malate dehydrogenase, MDH; somatic emb1yos arising from the nucel­ isocitrate dehydrogenase, IDH; phospho­ lus. Competition between these different glucose isomerase, PGI; aspartate amino emb1yos results in partial apomixis. The transferase, AAT), as described by apomixis rate is dependent on the geno­ OLLITRAULT et al. (1992). Four of them are type and environmental conditions (KHAN unlinked (LuRo et al., 1995 and our and ROOSE, 1988). Moreover, there can be unpublished data), so more than 94% of a low percentage of polyploicls clue to the zygotic seedlings could be detected. fertilization of non-reduced ovules or spontaneous polyploiclization of nucellar The ploidy level of each seedling was tissues (ESEN and SOOST, 1977; lWAMASA evaluated by flow cytomet1y, as described and NITO, 1989). by OLJ.JTRAULT and MICHAUX-FERmERE (1992). Samples were analysed at the Citrus volkameriana is a very vigorous INSERM Unit 291 laborato1y in and highly heterozygous rootstock, as Montpellier under the supervision of shown by isozyme analysis (OLLITRAULT et C. Duperray. al., 1992). It was therefore predicted that zygotic seedlings would show ve1y mar­ Vegetative seedling vigor was estimated 1 year after planting according to three ked diversity. However, C. volkameriana seedling populations are generally bimo­ characters: seedling height, number of dal, i.e. a highly vigorous true-to-type internodes and diameter of the second internode. A synthetic index of vigor was population and, conversely, a very weak population. established from these characters by prin­ cipal component analysis. In the present paper, we have analysed the genetic status (somatic or zygotic ori­ gin, and ploidy level) of C. volkameriana seedlings and relationships with plant results vigor. Twenty-nine zygotic plants were identi­ fied by isozyme analysis (Photo 1), while

flow cytometry analysis revealed two tet­ raploid plants (Fig. 1). These two plants, material and methods which had enzymatic patterns identical to The genetic origins of 135 Citrus volka­ C. volkameriana, certainly arose through meriana greenhouse-grown seedlings at spontaneous tetraploidization of somatic the SRA San Giuliano station were ana­ cells. The enzymatic patterns of the zygo­ lysed. tic seedlings suggested that they mainly

398 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• GENETICS / BIOTECHNOLOGY e

•• •

Photo 1 C. Volkameriana (last sample) exhibits five heterozygous loci for the 4 analysed enzyme systems. From the combination of these sys tems it appears that samples n°1, 2, 3, 4, 5, 6, 7, 8, 9, 13 and 14 are zygotic seedlings. The others samples are considered as nucel/ars with a 3% risk (if the five MDH heterozygous loci are unlinked). arose via self pollination. The enzymatic patterns and flow cytometry histograms of the 104 other plants were identical to C. volkameriana; they were thus consid­ GO/G 1 Control GO/Gl Sample ered to be somatic diploids of nucellar origin. The three morphological characters stud­ ied were very highly correlated, and the vigor index established from the first axis in the principal component analysis rep­ resented more than 90% of the total vari­ ance. This vigor index was used to estab­ lish the vigor distribution between the three genetic groups (Fig. 2). The vegeta­ tive vigor levels of most of the zygotic seedlings (23/29) were substantially lower than those of plantlets of nucellar origin and, similarly, tetraploid plantlets showed G2/M Control very depressed development (Table 1).

C Volkamerlana (Tetraploid) discussion The two main vigor classes in C. volka­ meriana seedlings could be associated with different genetic origins. The weak GO/G l Sample GO/G 1 Control plants were diploid zygotic or somatic tet­ raploids, while the vigorous plants were mainly nucellar diploids.

Inbreeding was previously reported for G 2/M Controls other polyembryonic Citrus and Poncirus species (KHAN and ROOSE, 1988; SOOST C Volkamerlana (Diploid) and CAMERON, 1975). Apomixis certainly Relative DNA content caused deleterious recessive mutations with respect to the heterozygous status, as noted in self-fertilized zygotic seed­ Figure 1 lings. The high proportion of depressed Fl ow cytometry analysis of ploidy levels in C. volkameriana seedlings. C. volkameriana seedlings suggested the Tahiti lime (trip/aid) was used as internal control.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 399 •• • OLLITRAULT and JACQUEMOND

(>20%). This could indicate that the rate Table 1 of spontaneous tetraploidization of nucel­ Numbers and characteristics of nucellar, zygotic and tetraploid lar tissues was higher than that evaluated seedlings. for tetraploid plantlets. Indeed, the emer­ gence of zygotic seedlings must be favou­ Number DI NI H red by the ve1y low vigor of tetraploid CV CV Mean CV of plants % Mean Mean emb1yos. •

Nucellars 104 770 11.6 013 72.5 0.11 140 0.17 Zygotics 29 21.5 6.6 0.36 22.7 0.59 51 089

Tecraploids 2 1.5 5.6 0.20 24.5 0.10 30 0.04

Dl: Diameter of the second internode (mm) Nl: Number of internodes H: Hight of the seedling (cm) references ESEN A., SOOST R.K., 1977. Relation of unexpected polyploids to diploid megagametophytes and embryo: endosperm 20 ploidy ratios in Citrus. In: I. Congreso mun­ dial de Citricultura, Murcia, Valencia, 18 29 April-10 May 1973. Vol. II, 0. Carpena ed., International Society of Citriculture, 16 p. 53-63. 14 IWAMASA M., NITO N., LING J.T., 1989. 12 Intra- and intergeneric hybridization in the E orange subfamily, Aurantioideae. In: Proc. 10 6th lnt. Citr. Cong. Middle-East, Tel Aviv, Israel, 6-11 March 1988. R. Goren and 8 K. Mendel eds., Margraf Scientific Publis­ hers, Germany, p. 123-130. 6 KHAN I.A., ROOSE M.L., 1988. 4 Frequency and characteristics of nucellar and zygotic seedlings in three cultivars of trifo­ 2 liate orange. J. Amer. Soc. Hort. Sci. 113 (1): 105-110. 0 1 •• Vigor index LURO F., LORIEUX M., LAIGRET F., BOVE J.M., OLLITRAULT P., 1995. Tetraploids • Zygotics • Nucellars Genetic Mapping of an lntergeneric Citrus Hybrid using Molecular Markers. Fruits, vol. Figure 2 49 (5-6): 402-406. Vigor indexes for presence of several unlinked mutations of C. volkameriana seedlings this kind. However, the morphology of OLLITRAULT P., FAURE X., NORMAND F., 1992. of three genetic origins. Citrus rootstock characterization with bark some zygotic seedlings was very similar and leaf isozymes; application for distingui­ to the nuceUar morphology. For agro­ shing nucellar from zygotic trees. In: Proc. of nomic or varietal tests, it is thus necessary VII lnt. Citrus Cong., Acireale, Italy, 3-8 March to certify the nucellar origin of seedlings 1992. lnt. Soc. of Citriculture, p. 338-341. by molecular analysis (OLLITRA LT et al., 1992). OLLITRAULT P., MICHAUX-FERRIERE N., 1992. Application of flow cytometry for Citrus gene­ The low proportion and vigor of sponta­ tics and breeding. In: Proc. VII lnt. Citrus neous C. uolkameriana tetraploids was in Congr., Acireale, Italy, 3-8 March 1992. lnt. agreement with results obtained in other Soc. of Citriculture, p. 193-198. Citrus species (IWAMASA and NITO, 1989). SOOST R.K., CAMERON J.W., 1975. Despite inbreeding, the proportion of Citrus. In: Advances in Fruit Breeding, Purdue zygotic seedlings was relatively high University Press, Indiana, p. 507-540.

400 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards Protoplast Fusion in Citrus

P. 0LLITRAULT, D. DAMBIER •••••••••••••••••••••••••••••••••••••••••••••• CIRAD-FLHOR, BP 5035 Protop/ast fusion, a prelude for regeneration of interspecific and 34032 Montpellier cedex 01 France intergeneric somatic hybrids and cybrids, is a promising area of research

C. CABASSON, C. TEISSON for the improvement of Citrus rootstock and cultivars. (2) CIRAD-BIOTROP, BP 5035 34032 Montpellier cedex 01 •••••••••••••••••••••••••••••••••••••••••••••• France

F. LURO SRA San Giuliano INRA/CIRAD-FLHOR 20230 San Nicolao introduction obtaining intergeneric Corsica Protoplast fusion is an increasingly impor­ somatic hybrids tant component of citrus breeding pro­ ° Research conducted in France over the Fruits, vol. 49, n 5-6 grammes (OLLJTRAULT and LURO, 1995). It 2 p. 401-403 (English) past 5 years by CIRAD-FLHOR , CIRAD­ enables breeders to overcome some 3 4 p. 481-482 (French) BI0TROP and INRA (agricultural constraints linked with sexual hybridiza­ research station at San Giuliano, Corsica) tion and is applied to develop new root­ on embryogenesis in citrus materialized in stocks and cultivars for high quality, sus­ r 1994 when the first Citrus reticulata tainable and environment-f iendly citrus (mandarin) + Fortunellaja ponica (kum­ cropping. quat) intergeneric hybrids were obtained Protoplast fusion can be used in rootstock in the CIRAD-BIOTROP laboratories. improvement programmes to accumulate About a hundred plants were thus rege­ reticulata genes with resistance or tolerance to bio­ nerated after electrofusion of C. tic and abiotic factors, irrespective of the emb1yogenic callus protoplasts and F.ja ponica heterozygosis level of the parents. leaf protoplasts (Photos 1-3). Flow cytometry was used to determine In addition, there is a very broad range of ploidy levels (Fig. 1), and isoenzyme ana­ genetic resources that can be utilized r lysis confirmed their hybrid status. These since genomes f om sexually incompa­ tetraploid hybrids coud be sexually hybri­ tible species and genera can be combined dized with diploid cultivars to introgress by this fusion technique (GMITTER et al., flavours, lateness, cold tolerance and 1992). (1) CIRAD: Centre kumquat bacterial canker resistance in de cooperation internationale Protoplast fusion also shows promise for mandarin-type triploid cultivars. en recherche agronomique diversification programmes through the pour le developpement. development of triploid cultivars, espe­ (2) CIRAD-FLHOR: cybrid identification cially in mandarin (OLLITRAUJ:r et al., Departement des productions As part of a cooperative project with fruitieres et horticoles. 1995). The pool of tetraploid progenitors, IFAS5 (Dr. J. Grosser, Florida, USA), (3) CJRAD-BIOTROP used in sexual hybridization with diploids Laboratoire des biotechnologies to produce triploid cultivars, could be diploid cybrids (nuclear genome from appliquees al'amelioration des enriched considerably. one species combined with cytoplasmic plantes tropicales. genomes of another species) were identi­ (4) !NRA lnstitut national Fertility has also been restored in somatic fied by two techniques: flow cytomet1y de la recherche agronomique. hybrids obtained from sterile elite culti­ and RFLP marker analysis of nuclei and (5) !FAS: Institute of Food vars (OHGAWARA et al., 1991). Results cytoplasm (Photos 4 & 5) from plants 1 4 and Agricultural Sciences obtained by CIRAD and INRA in this regenerated following fusions between (University of Florida). area are presented. C. deliciosa (mandarin) callus protoplasts

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 401 • • OLLITRAULT et al.

and C. sinensis (sweet orange) leaf proto­ prospects plasts, as well as between C. deliciosa cal­ lus protoplasts and C. paradisi (grape­ Somatic hybridizations of various inter­ fruit) leaf protoplasts (LuRo et al., 1995). specific [ C. reticulata + C. sinensis; C. reticulata + C. paradisi; C. reticulata The studies highlighted the systematic + C. lemon (lemon); C. reticulata + presence of nonrecombined mitochon­ C. aurantifolia (lime)] and intergeneric drial and chloroplast genomes from the ( C. reticulata + Poncirns trifoliata and C. deliciosa parent (embryogenic calli) C. aurantium + Eremocitrns glauca) com­ associated with nuclear genomes from binations have been carried out and C. deliciosa, C. sinensis or C. paradisi. embryos are currently being propagated. The results were similar to those obtained These hybridizations are to be used in by TusA et al. (1990) and SAITO et al. rootstock and cultivar improvement pro­ (1993). They suggest that Citrns somatic grammes and should also provide valua­ embryogenesis regeneration capacity is ble information on polyploid genetics . under cytoplasmic control. Cybrids obtai­ ned in this way are ideal for assessing the Fusions between diploid protoplasts and Figure 1 involvement of cytoplasmic genomes and haploid protoplasts (derived from haploid Analysis of ploidy levels nucleus-cytoplasm interactions in determi­ plantlets obtained by induced gynogene­ of plants regenerated after ning phenotypic characters. sis) will soon be performed in order to somatic fusion between synthesize triploids directly. The origins C. deliciosa and F. japonica. of plants regenerated in these different programmes will be checked with mole­ cular markers and by flow cytometry. All Gl plant material obtained by protoplast control 2x 3x fusion will then be introduced in the CIRAD-FLHOR/INRA multilocation experi­ mental network to assess their potential under Mediterranean and tropical condi­ tions and' their tolerance to the main control diseases affecting such crops. • 3x acknowledgements We thank Mr. R. Haicourt and Mr. D. Sihacharkr for their valuable advice concerningthe protoplast Relative size of the genome electrofusion technique. C. deliciosa F Japonica . references GMITIER F.G., GROSSER J.W., MOORE G.A., 1992. Citrus. In: Biotechnology of perennial fruit crops, Hammmerschlag and Litz ed., CAB International, p. 335-369. control 3x LURO F., CABASSON C., GROSSER J., OLLITRAULT P., 1995. Utilisation des marqueurs RFLP et de la cytometrie en flux pour !'analyse genetique des plantules regenerees apres fusion de Tangequat protoplastes d'agrumes. In: Symposium mediterraneen sur mandarines, San Giuliano, France, 5-11 March 1995, (abstracts). OHGAWARA T., KOBAYASHI S., ISHII S., YOSHINAGA K., OIYAMA I., 1991. Fertile fruit trees obtained by somatic hybri­ dization: Navel orange (C. sinensis) and Relative size of the genome Troyer citrange (C. sinensis x Poncirus trifoliata) . Theor. Appl. Genet. 81: 141-143.

402 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY e

OLLITRAULT P., LURO F., 1995. Amelioration des agrumes et biotechnologie. Fruits (to be published).

OLLITRAULT P., F. LURO, ALLENT V., DAMBIER D., 1995. Diversification des mandariniers : apport des biotechnologies pour la creation de cultivars triplo"ides aspermes. In: Symposium mediter­ raneen sur mandarines. San Giuliano, France, 5-11 March 1995, (abstracts).

SAITO W., OHGAWARA T., SHIMIZU J., ISHII S., KOBAYASHI, 1993. Citrus cybrid regeneration fo llowing cell fusion between nucellar cells and mesophyll cells. Plant Science 88: 195-201.

TUSA N., GROSSER J.W., GMlnER F.G., 1990. Plant regeneration of "Valencia" sweet orange, "Femminello" lemon, and the inter­ specific somatic hybrid following protoplast fusion. J. Amer. Soc. Sci. 116: 1043-1046. Photo 3 Somatic hybrids between •••• C. deliciosa and F. japonica. I

•••� •··· ...... 2 3 4 5 6 7 8 9 10 11... 12 13 14 15 16 17 ....18 19 20 21 \

Photo 4 • Checking the origins of nuclei 3 4 5 -- 6 7 8 9 10 11 12 13 from plants regenerated after Photo 1 1 2 . protoplast fusion between Mixture of Citrus deliciosa embryogenic callus C. deliciosa and C. sinensis. protoplasts and Fortunella japonica leaf protoplasts. 1: C. deliciosa 2: C. sinensis. Plants regenerated afterfu sion: • 3-10 and 12-15: cybrids ... 11 and 16 to the end: plants regenerated from unfused .. • C. deliciosa protoplasts . -

Photo 5 Checking the origins of cytoplasm (mitochondrial probe) from plants regenerated afterprotoplast fusion between C. deliciosa and C. sinensis. Photo 2 1: C. deliciosa Proembryos 28 days afterelectrof usion 2: C. sinensis of C. deliciosa and F. japonica protoplasts. 3-13: plants regenerated afterfu sion.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 403 Genetic Mapping of an lntergeneric Citrus Hybrid Using Molecular Markers

F. LURO •••••••••••••••••••••••••••••••••••••••••••••• SRA San Giuliano A Citrus genetic linkage map was constructed: it can be used INRA/CIRAD-FLHOR in interspecific and intergenetic hybrid heredityst udies and for determining 20230 San Nicolao Corsica genes of agronomic interest. M. LORIEUX •••••••••••••••••••••••••••••••••••••••••••••• Laboratoire AGETROP GIRAD, BP 5035 introduction gation distortions. The different analytical 34032 Montpellier cedex 01 procedures are discussed in terms of the France Many problems are encountered when underlying biological models. F. lAIGRET conducting genetic analyses of citrus J.M. Bove hybrids, mainly for the following reasons: Laboratoire de biologie (1) high heterozygosity, (2) partial apog­ material and methods cellulaire et moleculaire amy (polyembryonic seeds with embryos hybridization INRA Bordeaux of nucellar origin and an emb1yo of zygo­ BP 81 The progeny investigated (52 plants) 33883 Villenave-d'Ornon cedex tic origin), (3) the high numbers of prog­ were obtained by crossing a Citrus gran­ France enies, and (4) the long juvenile phase (GMITTER et al., 1992). The genetic dis (cv Seedless pummelo) fe male parent P. 0LLITRAULT with an intergeneric hybrid male parent. CIRAD-FLHOR, BP 5035 improvement potential of cultivated citrus 34032 Montpellier cedex 01 has been enhanced through biotechno­ The later was obtained through a Citrus France logical advances and the development of resh ni Hort. ex Tan. (cv Cleopatra man­ molecular marker techniques. Gene map­ darin) x Po ncirus trifoliata L. Raf. ping has been carried out by several (cv Swingle) cross. laboratories (DURHAM et al., 1992; JARREL et Fruits, vol. 49, n°5-6 al., 1992). markers p. 404-408 (English) About one hundred markers were obtai­ p. 483-485 (French) A citrus g,-:netic linkage map was con­ structed using different types of molecular ned. About half of these were RFLP mark­ markers (isoenzymes, RFLP and RA PD), ers from two cDNA banks (LuRo, 1993). r and the main applications are: Forty probes were derived f om a cDNA - investigate chromosomal heredity in bank that we obtained from cv Valencia interspecific and intergeneric hybrids, Late orange (Citrus sinensis) mRNA. Four - determine the number and assess the other probes from a Citrus jambhiri bank impact of detrimental or unsuitable muta­ were supplied by Dr. Roose of Riverside tions when crossing polyemb1yonic spe­ University, California (JARRELL et al., 1992). cies, Only cDNas with slightly repeated - locate genes involved in the develop­ sequences or unique sequences showing ment of agronomically interesting charac­ polymorphism between the parents used ters. Markers associated with these genes in the cross were considered. could then be used for early selection of About forty amplified fragments segrega­ progenies and possible progenitors for ting in the progeny were selected with a breeding programmes (DE VrENNE, 1984; system involving RAPD markers (WILLIAMS STUBER, 1989). Moreover, markers could et al., 1990) using operon primers. potentially be used to isolate and transfer Selection was based on the repetability of genes by genetic engineering (GANAL et the amplifications and intensities of the al., 1991 ). r amplified f agments. Heredity was also A linkage map of a citrus genome from a studied in a few fragments amplified hybrid obtained by a three-way cross is by primers representing microsatellite se­ presented, focusing specifically on segre- quences ([TCC]S; [GACA]4).

404 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY e

Seven isoenzyrne systems were utilized: deduced from the genetic differences bet­ isocitrate dehydrogenase (ICD), rnalate ween parents used in the cross (Fig. 1). dehydrogenase (MDH), phosphoglucose This progeny could thus be analysed for isomerase (PGI), endopeptidase (GOD all markers as that of a test cross. and phosphoglucomutase (PGM). Three Forty-six fragments specific to the inter­ isoenzymatic loci (LAP, GOT and PGI) generic hybrid (19 mandarin and were heterozygotic in pummelo and six 27 Po ncirus) and 24 fragments specific to (End, PGM2, LAP, GOT, ICD and MDHl) the maternal genome (pummelo) were in the intergeneric hybrid (male). amplified by 25 primers using the RAPD genetic mapping technique along with marker segregation in the progeny (Fig. 2). The genetic linkage maps were con­ structed using the Mapmaker software The genetic linkage map constructed by package (LANDER et al., 1987). Several Lod the Mapmaker program included 95 markers distributed in 12 linkage score (logarithm of odds ratios) thresholds groups (with a basic chromosome num­ were tested to construct linkage groups. ber of 9) (Fig. 3). Nine loci did not HALDANE's (1919) genetic mapping func­ belong to any linkage group. The map tion was used. The recombination fre­ was 1503 cM long (Haldane function), quencies were calculated with the which is about two-thirds the total length Mapmaker program. of the genome QARRELL et al., 1992). The The linkages and recombination frequen­ groups were established with a Lod score cies (r) for each pair of markers were also > 3 and maximum recombination fre­ calculated using the Genepop software quency of 0.3 (corresponding to the program (OLLITRAULT, 1987). This program approximate maximal detectable fre­ measures linkages by the chi-square test quency for a backcross of 52 plants). of independence (MATHER, 1957). This test takes segregation distortions into consid­ segregation distortions eration relative to calculated theoretical r Thirty-nine of the 104 markers mapped f equencies for each class; it is based on (37.5%) showed significant (chi2 at 5% marginal frequencies of the contingency level) distortion as compared to the theo­ Figure 1 table. retical 1/1 segregation. In contrast, in RFLP profile. Results pummelo, only 3 of 45 markers (6.5%) of hybridization between DNA results and discussion segregated in a non-Mendelian manner. of parents and their progeny, Abnormal marker segregations have digested by EcoRV and genetic linkage map the cDNA probe 0.30. already been observed in citrus, espe­ 1: cv Seedless pummelo In all cases and for each individual cially in hybridizations of plants of differ­ 2: cv Cleopatra mandarin; hybrid, genetic compositions of male and ent genera (TORRES et al., 1985; OLLITRAULT 3: Poncirus trifoliata; female gametes for the intergeneric & FAURE, 1992; DURHAM et al., 1992; JARRELL 4: FA Q 30573 hybrid C C. reshni x P. trifoliata) were et al, 1992) (Cleopatra mandarin x Poncirus trifoliata).

Progeny --..:!t======-----���=�===::=::======:==I>' •

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 405 • • • LURO et al.

Progeny

Figure 2 Segregation of RAPO markers. Note that linkage group 1 on the interge­ intergeneric hybrid than in the female Results of amplification with neric hybrid genetic linkage map was grapefruit parent, were in line with the primer OPK4, ONA fr om constructed solely with markers showing biological data: cv Seedless pummelo (1) segregation distortion, including almost x FA Q 30573 (4) hybrids. • data from the literature demonstrate half of all loci affected. Detailed analysis The two 590 bp and 570 bp that distortions are much more common of distortions in linkage group 1 revealed fragments were transmitted in interspecific and intergeneric hybridiza­ by cv Cleopatra mandarin (2). that all of the markers had undergone tions than in intraspecific hybridizations Lane 3: Poncirus trifoliata. selection in favour of the cv Cleopatra (structural heterozygosity definitely has an parent. Two hypotheses could be put important impact). The overall selection forth: in linkage group 1 could be clue to such - this is not a real group because the dis­ tortions caused false linkages; structures; - this group actually corresponds to a • the two intergeneric hybrid parents of chromosome and thus the widespread the intergeneric hybrid were polyem­ distortion is probably partially clue to bryonic whereas the pummelo was structural heterozygosity of this chromo­ monoembryonic. It is therefore quite some set. likely that the genetic load was lower in It is quite unlikely that there was a false the pummelo parent. Indeed, facultative linkage group since the chi2 tests for apogamy promotes builclup of mutations, inclepenclence, to assess possible linkages in the heterozygotic state, that are detri­ for each pair of markers, were always sig­ mental or unsuitable in the homozygotic nificant. Counter-selection of a whole state. These mutations can induce abnor­ chromosome could be the result of struc­ mal segregations during crossing (gametic tural heterozygosity and gametic selection selection) and selfing (gametic and zygo­ favouring one or several cv Cleopatra tic selections) (LuRo et al., 1995). The mandarin genes. At one of its ends, link­ presence of structural heterozygosity can age group 3 incluclecl five markers show­ extend the effect of a locus selected to ing segregation distortion. A unimoclal large chromosomal fragments or entire distortion gradient was obtained, which chromosomes. In such cases, the deter­ peaked around markers Vlc3.18 and mined linkages can only be confirmed by VLc0.37. This pattern seems to indicate comparisons with a map without any seg­ the presence of a single gene that was regation distortions; selected in the vicinity of these two mark­ • due to pollen competition, it is quite ers. likely that segregation distortions occur­ The very marked differences, obtained for ring in gametes are mainly the result of molecular marker segregation distortion male parent input rather than that of the levels that were much higher in the male fe male parent.

406 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • GENETICS / BIOTECHNOLOGY •

7(163 cM>

2(161 cM) S (222 cM) 6(52 cM) RLl4b 3(126cMlI. ITT (20.3 )- (22.7 )_ 3.21 7.2 0.2A OPP6b 0.12GACA GOT rDNA PU8a /.l OPP2 PI.lb PLJO. PKl71 (33.8 ) - IU (8.8 ) - l.2S (15.4)- PK17c O.U1 OP2a • PLl4a (:!5.3 l OPF7a (12.4)- ( 0.0 l - • PLI lb ( 0.0 ) 1.22b MDHl 1.6 OPll> • ( 9.6 l 11.l 120.l l- 1.19 OP3a 112.9)- PL18d lB . OPlb 3.08 0.13 (17.4)- PK17b 2.6 I.Oil 3.02a \.)6 (23.l )- .. 12A.l l- (13.SJ- ... (22.6 )- 14.7 OPla J.8 ... s., 2.03 3.18 0.1 lc 10.S0{) ( )5.ll - I.I ICD 0.37 0.8 I 2.3 l 3.09 Pl18b • 3.29 0.04 133.8 )- 120.1 1- 0.36 TCC- IS.3 O.l lb 9 (2S3 cM) .. OPlc 4(51 cM) 0Pl'7b 29.0 (19.2 ) - PL18c 8151 cM) 0.18 .J ) (19.8 )- PK4b (30 - PK2a (21.2 ) - ()8.1 )- l''" J.Oll>PPl4a (29.4)- I (169 cM) (20.2 )- I.I RLl41 1.26b PK6b ... t PU·· · ()0.3 )- "" 0.0Sa •••• ( 2.0 ) 0.02 ( 0.0 l OD I 0.0 ) End3.48 Jl.3 (25.0 l

/' 2.( I PG2 13.S 11 5.4 l 10(1 l4cM) Pll la 11 ll09cMIl'L2a RL32 < 1.9 l OPF6a ••.....••• OPP4b • 2.02 ll.l00 0.30 ..... 16.� 00]) 1.OPF1c0l ••• 126.S i- (13.Sl - . PF14h ••••• (JO.I )- IS.9 17.1 (15.6) - 17.9, PGJ ..... 3.07b .. l!i.4 ..... PK2b 3.071 RLOJ I 9.6 ) (23.4 )- 0.09 .io.o (24.7( 6.1 l} ll 132 cMl PK4a 0.05b C:D.3 >- (19.9 ) - 1 SS)- PK4c J.14 (28.8 )- PL16b (3.4 ) JJ.O (:?7.5 )- RU4c " (961- {5.7 ) LAP (29.7 )- 3.05• 00 t �.s , r • 00 PKI 2 ... PK4d 00 2.01 ... 11311- 0.31 OPJb PL18c Figure 3 Genetic linkage map fo r the conclusion molecular markers for selection in plant intergeneric hybrid (C itrus reshni breeding programmes. Chromosome inver­ x Poncirus trifoliataJ. Numbers Construction of a genetic linkage map left of the linkage groups indicate sions and the presence of false linkages the recombination frequencies from molecular markers is the first step have much more negative effects than sim­ towards fully understanding the underlying calculated with the Mapmaker ple distance estimate biases when attempt­ software package, the marker mechanisms involved in citrus reproduc­ ing to locate genes or QTLs. Moreover, names are listed on the right. tion. It also provides a means to locate introgression of a chromosome fragment, Asterisks indicate markers genes of agronomic interest, e.g. showing promoted by a molecular marker with a with segregation distortion: disease resistance (tristeza, citrus bacterial distortion effect, leading to a false linkage *· chi2 significance at 5% canker, phytophtora, etc.). A partial gene­ with a sought-after trait, will result in **· chi2 significance at 1% tic map containing one hundred markers breeding failures. It is thus essential to ***· chi2 significance at 0.1% 2 divided into 12 linkage groups is still insuf­ gain access to genetic information for the ****· chi significance at 0.01% *****: chi2 significance at 0.001 %. ficient and should be supplemented with studied species in order to pinpoint the new markers. However, the presence of origins of distortions. This will facilitate segregation distortions in specific genomic choices of suitable assessment methods for regions could jeopardize the use of these future genetic mapping programmes. e

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 407 •• • • LURO et al.

references

DE VIENNE D., 1984. MATHER K., 1957. Limites et perspectives des marqueurs mole­ The measurement of linkage in heredity. culaires. Le selectionneur franc;:ais33: 35-46. London, UK, Methuen & Co, 149 p.

DURHAM R.E., LIOU P.C., GMITIER F.G., OLLITRAULT P., 1987. MOORE G.A., 1992. Evaluation genetique des sorghos cultives Linkage of restriction fragment length poly­ (Sorghum bicolor L. Moench) par !'analyse morphisms and isozymes in Citrus. Theor. conjointe des diversites enzymatique et mor­ Appl. Genet. 84: 39-48. phophysiologique. Relation avec les sorghos sauvages. PhD Thesis, Universite de Paris XI, GMITIER F.G., GROSSE R.J.W., MOORE G.A., Orsay, France, 187 p. 1992. Citrus. In: Biotechnology of perennial fruit OLLITRAULT P., FAURE X., 1992. crops. Hammerschlag et Litz. CAB Interna­ Systeme de reproduction et organisation de la tional, pp 335-369; diversite genetique dans le genre Citrus. In: Colloque international: Complexe d'especes, GANAL M.W., BONIERBALE M.W., ROEDER M.S., flux de genes et ressources genetiques des PARK W.D., TANKSLEY S.D., 1991. plantes. Paris, France, 8-10 January 1992. Genetic and physical mapping of the patatin Paris, France, Lavoirier, Bureau des res­ genes in potato and tomato. Mol. Gen. sources genetiques, p. 135-151. Genet. 225: 501-509. PERRIER X., OLLITRAULT P., DUBOIS C., 1992. HALDANE J., 1919. Cartographie du genome et QTL's. The combination of linkage values and the L'approche biometrique et les contraintes calculation of distance between the loci of biologiques. Fruits, Special issue on Citrus linked factors. J. Genet. 8: 299-309. 47: 135-144.

JARRELL D.C., ROOSE M.L., TRAUGH S.N., STUBER C.W., 1989. KUPPER R.S., 1992. Marker based selection for quantitative traits. A genetic map of Citrus based on the segre­ Vortrage fur Pflanzenzuchtung 16: 31-49. gation of isozymes and RFLPs in an interge­ neric cross. Theor. Appl. Genet. 84: 49-56. TORRES A.M., MAU-LASTOVICKA T., WILLIAMS T.E., SOOST R.K., 1985. LANDER E.S., GREEN P., ABRAHAMSON J., Segregation distorsion and linkages of Citrus BARLOW A., DALY M.J., LINCOLN S.E., and Poncirus isozymes genes. J. Hered. 76: NEWBURG L., 1987. 289-294. Mapmaker: An interactive computer package for constructing primary genetic linkage maps WILLIAMS J.G.K., KUBELIK A.R., LIVAK K.J., of experimental and natural populations. RAFALSKI J.A., TINGEY S.V., 1990. Genomics 1: 174-181. DNA polymorphisms amplified by arbitrary pri­ mers are useful as genetic markers. Nucleic LURO F., 1993. Acids Res. 18: 6531-6535. Utilisation des marqueurs moleculaires pour la cartographie du genome et les etudes genetiques chez les agrumes. PhD, Thesis, University of Bordeaux II, France, 187 p.

LURO F., LORIEUX M., LAIGRET F., BOVE J.M., • ••• OLLITRAULT P., 1995. Cartographie du genome des agrumes a l'aide des marqueurs moleculaires et distor­ sions de segregation. In: Les colloques de l'INRA: Techniques et utilisations des mar­ queurs moleculaires. Montpellier, France, 29- 31 March 1994. Paris, France, INRA, n· 72, p. 69-82.

408 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• IV. Crop protection

VIROLOGY Tristeza Survey in the West Indies. c. BuJAooux et al. [Tristeza of citrus, identification, biological control, vectors, Aphididae, parasitoids, Caribbean] 410-414 [Tristeza des agrumes, identification, Jutte biologique, vecteur de maladie, Aphididae, parasito'ide, Cara'ibes] [Tristeza de los agrios, identificaci6n, control biol6gico, vectores, Aphididae, parasitoides, Caribe]

PHYTOPATHOLOGY Serious Citrus Dieback in Colombia Caused by Ceratocystis fimbriata. x. MouR1cHoN r [citrus f uits, Ceratocystis, identification, fu ngal diseases, symptoms, lesions, Colombia] 415-416 [agrume, Ceratocyst is, identification, maladie fongique, symptome, lesion, Colombie] [frutas citricas, Ceratocystis, identificaci6n, enferrnedades fungosas, sintomas, lesiones, Colombia]

ENTOMOLOGY Research and Control Programmes Against Fruit Flies in Reunion. s. Qu1uc1 [Tep hritidae, Ceratitis. Bactrocera . .J 417-420 [ ... biology, ecology, insect control, biological control, oviposition, trapping, fruit trees, Reunion] [. .. biologie, ecologie, Jutte antiinsecte, Jutte biologique, ponte, piegeage des animaux, arbre fruitier, Reunion] [ ... biologia, ecologia, control de insectos, control biol6gico, oviposici6n, caza con trampa, arboles frutales, Reunion]

The Fruit Fly Research Programme in New Caledonia. J.M. LEMONTEY et al. [Tepbritidae. Bactrocera, Pnmus persica. Ps idium guajava .. .] 421-427 [. .. population distribution, disease surveillance, infestation, animal husbandry, postharvest control, New Caledonia] [ ... distribution des populations, surveillance epidemiologique, infestation, elevage, Jutte apres-recolte, Nouvelle-Caledonie] [ ... distribuci6n de la poblacion, vigilancia de enfennedades, infestaci6n, ganader\a, control de plagas postcosecha, Nueva Caledonia]

Inventory of Insect Fauna Specific to Cultivated Fruit Trees of Northern Cote d'Ivoire. K. N'GurnA [Heteroptera, Coleoptera, Diptera, Hy menoptera, 011hoptera, Lepidoptera, Homoptera, Neurop tera, Dictyoptera ..] 428-429 [. .. insect, fruit trees, surveys, Cote d'Ivoire] [. .. insecte, arbre fruitier, enquete, Cote-d'Ivoire] L .. insect6, arboles frutales, encuestas, Cote d'Ivoire]

Inventory of Insect Fruit Pests in Northern Cote d'Ivoire. K. N'GurnA [Fruit damaging insects, citrus fruits, mangoes, guavas, papayas, cashews, Cote d'Ivoire) 430-431 [Insecte depredateur des fruits, agrume, mangue, goyave, papaye, pomme cl 'anacarde, Cote-d'Ivoire) [lnsectos depredadores de los frntos, frutas citricas, mango, guayaba, papaya, anacardo, Cote d'Ivoire]

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 409 • Tristeza Survey in the West Indies

C. 8UJADOUX •••••••••••••••••••••••••••••••••••••••••••••• M.L. CARUANA A comprehensive indexing operation has been under way since 1993 Laboratoire de virologie in experimental orchards of Martinique and neighbouring islands to map CIRAD-FLHOR BP 5035 the distribution of citrus tristeza virus in the West Indies. Behavioural 34032 Montpellier cedex 01 studies on vectors of this disease have been conducted for potential France biological control purposes. Fruits, vol. 49, n°5-6 •••••••••••••••••••••••••••••••••••••••••••••• p. 410-414 (English) p. 488-490 (French) introduction islands were then contaminated, with severe tristeza strains detected in Trinidad Citrus cropping is expanding in the West and Tobago, Jamaica, the Dominican Indies, mainly because of the increased Republic and even Cuba. demand for oranges to be processed for juice. Through its research, CIRAD­ CTV has a very high destructive potential FLHOR is striving to promote citrus pro­ in citrus orchards of the region since sour duction that is fully adapted to local orange is the most com.mon rootstock conditions. In 1988, an experimental and its associations with orange, manda­ orchard was thus set up at Riviere rin and grapefru it are very susceptible to Lezarde (Martinique) to study citrus beh�1- the virus. The CTV front should thus be viour in a wet tropical climate and propa­ carefully monitored. gate indexed plant material. An indexing operation was set up by However, there is a new citrus disease CIRAD-FLHOR at Fort-de-France (Martini­ threat in the region, as signaled by the que) in 1993 to deal with this situation. detection of the brown citrus aphid. The operation is focused on trees from To.xoptera citricid11s Kirkaldv. This is the SRA lines, which are distributed as virus­ r most efficient vector of citrus tristeza virus f ee material. Some private citrus orchards (CTV), which is transmitted semi-persis­ in Martinique and hundreds of neighbou­ tently. ring islands have also been surveyed. The overall aim of this operation is to assess The distribution range of T. cilricid11s the CTV status in Martinique and surroun­ r recently extended considerably f om its ding islands. This operation has been endemic zone in the southern hemisphere stepped up in conjunction with the towards North America (ROtSTACHER et al., increased brown citrus aphid (T.citrici­ 1991) dus) populations. ln Central America, the vector was first CTV control techniques mainly involve noted in Panama in 1985, then in Costa using resistant rootstock, cross protection Rica, Salvador and Honduras. and eradication. Eradication is used as a T. citridus colonized the West Indies preventive strategy if the virus has not yet within 5-6 years; it was first identified in been propagated. Control programmes Trinidad and Tobago in 1990, and then in have not seriously focused on the insect Martinique and Guadeloupe. vectors even though they represent the second most important means of trans­ In addition, CTV propagated and spread mission of CTV after grafting. to new areas along with its most efficient vector. A serious infection broke out in Studies have been carried out in the Central America in 1992. The Caribbean Mediterranean region on Lysiphlebus tes-

410 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / VIROLOGY e

taceipes Cresson (Hym., Braconidae), an 45 min and 1 h after the substrate was endoparasite of many different aphid spe­ plated. cies (STARY et al., 1988a and 1988b) • analysis of the results whose brown citrus aphid biocontrol potential looks promising (YOKOMI, 1992). The detection threshold was set 2- or This endoparasite should therefore be fur­ 3-fold higher than the mean optical ther identified and investigated in full densities (OD) of the healthy controls (X). detail. Samples with OD levels lower than 2X Studies on the behaviour of Toxoptera were negative (healthy), while those citricidus and Lysiphlehus testaceipes and higher than 3X were positive (contamina­ their interactions are also under way. ted). When the OD level was between these two levels, samples were tested a second time before determining the defi­ materials and methods nitive result. phytosanitary situation This detection threshold reduced the risks of making diagnostic errors for a disease plant sampling with o· tolerance. Sampling was done at two separate sites in the experimental Citrus orchard at insect rearing Riviere Lezarde (CIRAD-FLHOR, Marti­ aphids nique). Half of the trees at the first site, which includes 167 different varieties, T. citricidus aphids were reared on nur­ were tested. A quarter of the trees at the sery Mexican lime plants. The plants were second site, comprising trial plots with cut back regularly to promote the growth replicates of four trees, were sampled. In of shoots upon which the insect colonies surveyed private orchards of Martinique were reared. The aphids had been identi­ and neighbouring islands (Saint Lucia, fied and collected in the field. Saint Vincent, Grenada, Dominica, Aphids were reared in an insectproof Antigua, Nevis and Saint Kitts), samples cage under a shaded plastic tunnel in were collected from the most susceptible natural conditions: 25-35"C, 12L/12D trees: rootstock/susceptible variety asso­ photoperiod and about 95% relative ciations, or highly susceptible Citrus trees humidity during the rainy period. showing specific symptoms (e.g. vein The reared aphids were put in fine cotton clearing in lime). sleeves placed around pretreated young All samples were cut from young lignified shoots in cv Satsuma mandarin (Citrus shoots. unshiu) trees in the experimental CTV detection orchard. The progeny of these aphids were then put in separate sleeves. • ELISA-DAS technique These aphid were observed on a daily Indexing was performed using the ELISA­ basis to determine their characteristics DAS technique with polyclonal antibodies and behaviour under the environmental from SANOFI diagnostic kits. The tests conditions of Martinique. were carried out according to the manu­ facturer's instructions, and the conjugate parasitoids was depleted with 33 µl of healthy The L. testaceipes aphid endoparasites control plant material/ml conjugate solu­ were reared under the same conditions as tion. the aphids on Mexican lime branches col­ Mauritius papeda (Citrus hystrix) was pre­ lected from the T. citricidus rearing cages; ferentially used as control material for the these branches were replaced regularly. depletion operation since the virus puri­ Leaves with parasitized aphids (brown fied to develop the serum for the dia­ and mummy-like) were put in rearing gnostic kit had been obtained from this boxes and placed in a room at constant Citrus species. The ELISA plates were temperature (25"C) under a 12L/12D read on a spectrophotometer C 405 nm) photoperiod.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 411 • • 8UJAD0UX and CARUANA

L. testaceipes insects hatching on the same Martinique still seems to be CTV-free, but clay were grouped to determine the the survey carried out in the West Indies length of their adult life cycle. revealed the presence of the virus on two Two-day-old L. testaceipes adults from the of the seven monitored islands, i.e. rearing boxes were put in the cotton Antigua and, closer to Martinique, Saint f Lucia. Table 2 illustrates CTV propagation sleeves containing aphids of diferent patterns in the West Indies, which are ages (3-10 clays old) and densities closely correlated with the vector distribu­ (10-30 aphids). tions. These parasitoicls were removed after Although the sample sizes were small for 24 h with the host. After 4 clays, they these areas, CTV seemed to be wides­ were monitored daily to assess the length pread because of the broad distribution of their growth cycle and parasitism rates. range of its vector. It seems that the intro­ The fine cotton sleeves were placed on duction of CTV-infectecl plants was cv Navel orange trees. responsible for spreading the virus to the West Indies; it was then transmitted by T citricidus. results and discussion This hypothesis is supported by the fact indexing that in Martinique, where only SRA-type citrus material is present (originating from None of the 665 samples from Martinique the repository at San Giulano, Corsica, an were found to be contaminated with CTV island that is also virus-free), all ELISA (Table 1) results were negative. These results confirm the good health sta­ In contrast, CTV has been introduced on tus of the experimental orchard at Riviere neighbouring islands through buclwoocl Lezarcle (CTRAD-FLHOR, Martinique) or plants originating from Florida, a where many samples had been collected. region that has long been known to har­ The station can thus continue its plant bour tristeza. propagation activities under close super­ vision. The arrival of T citricidus should accele­ rate CTV transmission on islands that until No general conclusions can be drawn now have only been partially contamina­ from the present results because of the ted, and also boost between-island trans­ very low number of samples collected mission. A 24-h CTV watch is thus requi­ from Citrus growers' orchards (two were red on Martinique and Guadeloupe since surveyed); they only apply to the trees they both seem to be virus-free even after tested. T citricidus outbreaks.

Table 1 insect biology Results of CTV indexing analyses using the ELISA-DAS technique. aphids

Origin Total Trees Healthy Infested • life cycle length trees tested trees trees The mean life cycle length of T. citri­ r Martinique cidus, as determined f om a sample of Repository orchard 838 420 420 0 40 apterous aphids, was 17 days (± 2.47) Trial plots 956 220 420 0 under the environmental conditions of Plantations I 18 250 r- ) 25 0 Martinique (Fig. 1). Caribbean Saint Lucia • breeding rates 10 9 1 Saint Vincent 11 11 0 The mean number of offspring produced Grenada 2 8 8 0 by an apterous female aphid, as determi­ Dominica 18 18 0 Antigua 1 ned from a sample of 30 aphids, was 13 12 50 aphids (49.6 ±11.3), peaking at almost Nevis 11 11 0 Saint Kitts 7 7 0 60 aphids (59 ± 8) (Table 3). (1) estimated The first aphids were hatched after (2) no samples were found to be positive, despite observed CTV symptoms in the 6 clays, with maximums at 9-12 clays field. (Fig. 2).

412 • Fruits, vo l. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / VIROLOGY •

The high standard deviations for these results could be explained by variations Table 2 in the rearing system. Citrus production and variation in T. citricidus and CTV distributions in the Caribbean Basin (from AUBERT, 1992; CAO VAN, 1992, updated Nevertheless, the breeding efficiency in 1993). confirmed field observations, indicating that T. citricidus was widely distributed. It Citrus September 1991 September 1992 September 1993 seemed to be propagated at the expense Country output CTV T.citridus CTV T.citridus CTV T.citridus of Toxoptera aurantii, the main aphid MS SS MS SS S? present prior to the arrival of T. citricidus Costa Rica 34 + + + + + + + + but which is now rarely observed. Panama 86 + + + + + + + + + + + + + + + + The high parthenogenic breeding rates of Nicaragua 66 El Salvador 99 + - + + T. citricidus enabled it to quickly colonize Honduras 97 + + + the West Indies. It has now become the Guatemala 135 + - + most serious aphid citrus crop pest Belize 87 + - + + + because of the size of its colonies and the Mexico 3 081 Trin. & Tobago 16 + + + + + direct crop damage it causes. Saint Lucia + + + + + The geographic extension of this aphid is Martinique 2 Dominica 20 + also promoted by its marked ability to Guadeloupe 4. + + adapt to various climatic conditions; it is Puerto Rico 33 + + + + + found in wet tropical regions (Asia, South Dominican Rep. 84 + + + + + Jamaica 104 + ? + + + America, Africa south of the Sahara), and Cuba 898 (-) (-) + + in temperate areas (South America, Aus­ Grenada 2 + tralia, New Zealand (LECLANT et al., 1992). Saint Vincent 1 + Antigua + + parasitoids Nevis + Saint Kitts + • life cycle length In natural conditions, the mean life cycle CTV = citrus tristeza; MS = mild strain; SS = severe strain; S? = severiry unknown. length of L. testaceipes in the aphid is about 10 days (9.4 ± 1.46), as determined Figure 1 from a sample of 41 insects. Length of the T. citricidus life cycle. The mean length of the adult L. testa­ ceipes life cycle was found to be more 100 than 3 days (3.38 ± 0.9), as determined from a sample of 120 insects placed in 90 rearing boxes under unnatural conditions. 80 Parasitism rates assessed from field data were too heterogeneous to be interpre­ 70 ted. Nevertheless, it seems that L. testa­ 'J, 60 ceipes biocontrol could not alone reduce aphid populations to low enough levels 50 to hinder the spread of CTV. 40 conclusion ;:R0 30 The present results indicated that CTV 20 was not yet present in Martinique, despite the arrival of its vector T. citricidus at the 10 end of the 1980s. However, further in­ depth analyses should now be carried out 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 with more efficient techniques such as immunoprinting, PCR, etc. • Days

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 413 • • BUJADOUX and CARUANA

CAO VAN P., 1992. Table 3 Citrus Tristeza virus/Toxoptera citricidus: situation in Martinique, Guadeloupe (French Cumulated offspring for an apterous T. citricidus female. West Indies) and St Lucia. In: CTV and The results were calculated from the offspring of 30 female aphids. T. citricidus in Central America. Development of management strategies and use of bio­ Life cycle length (days) Total number of hatched aphids technology for control, September 14-19 12 34 7 (8.44)· 1992, Maracay, Venezuela, 287 p. CATIE 13 40 (9.17) (Costa Rica), University of Florida (USA), 14 45.5 (9 75) INIFAP-SARH (Mexico), Universidad Central de 15 49 6 (8 25) Venezuela, USDA (USA), p. 102. 16 53 2 (7.94) 17 57 6 (6 18) GILLETI J.M., MORRISSEY S.M., RAMSDELL D.C., 18 58.2 (5.76) 1986. 19 59 (8) Interpreting ELISA data and establishing the • (x) = standard deviation positive-negative threshold. Plant Disease 70 (8): 722-726.

LECLANT F., ETIENNE J., AUBERT B., 1992. 10 Alerte a la Tristeza en verger d'agrumes. Le puceron vecteur Toxoptera citricidus envahit 9 l'arc Cara·ibe. Phytoma (440): 32-34. 8 ROISTACHER C.N., GUMPF D.G., DODDS J.A., 7 LEE R.F., 1991. ..c 6 The threat of «the citrus killer». Citrograph 76 (10): 4-10. 5 "' STARY P., LYON J.P., LECLANT F. , 1988a. 4 Post-colonisation host range of Lysiphlebus "'C 3 testaceipes in the Mediterranean area (Hymenoptera, Aphidiidae). Acta Entomol. 2 Bohemoslov. 85: 1-11. 1 STARY P., LYON J.P., LECLANT F., 1988b. 0 Biocontrol of aphids by the introduced 1 2 3 4 5 6.I 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 11111 Lysiphlebus testaceipes (Cress.) (Hymenop­ Aphid ages (days) tera, Aphidiidae) in Mediterranean France. J. Appl. Ent. 105 : 74-87.

Figure 2 WALLACE J.M., 1978. T. citricidus breeding references The Tristeza disease complex. The citrus (mean number of aphids industry. University of California, USA. hatched/day/pare nt). AUBERT B., 1992. W. Reuther, E.C. Calavan and G.E. Carman, Le programme agrumes du CIRAD-FLHOR. eds, Volume IV, p. 87-109. Fruits, Special issue on Citrus 47: 99-102. YOKOMI R.K., 1992. AUBERT B., ETIENNE J., COTIIN R., LECLANT F., Potential for biological control of Toxoptera CAO VAN Ph., VUILLAUME C., JARAMILLO C., citricidus (Kirkaldy). In: CTV and T. citricidus BARBEAU G., 1992. in Central America. Development of manage­ Citrus Tristeza disease, a new threat for the ment strategies and use of biotechnology for Caribbean Basin. Report of a survey to control, September 14-19 1992, Maracay, Colombia, Dominican Republic, Guadeloupe, Venezuela, 287 p. CATIE (Costa Rica), Martinique and Trinidad. Fruits 47 (3): 393- University of Florida (USA), INIFAP-SARH 404. (Mexico), Universidad Central de Venezuela, USDA (USA), p. 194-198. CAMINADE J.M., 1992. Le virus de la Tristeza des agrumes (CTV): transmission par Toxoptera citricidus (Kirkaldy) (Hemipterae: Aphididae) en condi­ tions semi-contr61ees. Facteurs influenr;:ant la •••• concentration des plantes hates en parasites viraux. Thesis. Cergy-Pontoise, France, ISTOM, 77 p.

414 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards Serious Citrus Dieback in Colombia Caused by Ceratocystis fimbriata

tl> ••••••••••••••••••••••••••••••••••••••••••••• XAVIER MOURICHON CIRAD-FLHOR A new citrus disease in Colombia: symptoms and characteristics BP 5035 of the cultured fungus. 34032 Montpellier cedex 01 France • • • •••••••••••••••••••••••••••••••••••••••••• Fruits, vol. 49, n° 5-6 introduction This disease features internal spread of p. 415-416 (English) different colours of rot in parts of the p. 491-492 (French) A very worrisome canker disease was trunk or main branches or throughout recently detected in about 10% of citrus most of the ligneous tissues of the tree. crops in Colombia, but the actual damage Comparisons with other diseases can thus toll is probably even higher. Serious die­ only be made by analysing trunk or back has been observed over the last branch cross-sections. 3-4 years and threatens overall citrus pro­ duction in the Colombian coffee-growing Internal rotting quite often spreads from region. At the joint request of citrus grow­ the base of the trunk (above the graft) to ers in this region (Pereira, Manizales), the the top of the tree. Lesions are mainly Centro nacional de investigaciones de located in the stele; they then spread cen­ cafe (CENICAFE) and the Instituto colom­ trifugally in a black flame shape (espe­ biano agropecuario (ICA), a task force cially in lime, Photo 1), or much more was sent to assess the causes of this can­ extensively with a yellow serrated front. ker disease. External lesions are sometimes noted on symptoms various parts of the tree. The first disease symptoms (leaf wilt) often correspond to The greatest canker damage was noted widespread underlying internal cankers. on cv Tahiti lime (Photo 1) and orange trees (Photo 2). The initial symptoms appear on certain vegetative organs or diagnosis the whole tree; the leaves yellow and dry Photo 1 and 2 The fungus was isolated in vitro on vari­ up to various extents. The symptoms are Internal damage observed in ous media from samples collected at dif­ similar to those noted during Phyto­ parts of a citrus trunk or main ferent sites. Almost all cultures were deri­ branches. phthora attacks. ved from specimens cut from areas around lesions. Three genera of fungi were identified and their proportions evaluated: - Ce ratocystis sp. alone, 22% of the samples, - Ceratocystis sp. + Fusarium sp., 8%, - Fusarium sp. alone, 40%, - Diplodia sp., 30%. Morphological analysis of Ceratocystis in culture indicated that it belonged to C. fimhriata: a - typical perithecium and ascospore morphology, b conidial stage Chalara (endo- spores/ endoconidia), c - many chlamydospores (alieuspores) 1 2 (Photo 3).

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 415 e O MOURICHON

Microscopy analyses of ligneous tissue images in plum trees were published by fragments cut behind the growth front DEVAY et al. (1968). revealed many typical C. fi.mbriata chlam­ The problems encountered in isolating ydospores in the pitted vessels. However, this fu ngus in culture and the low isola­ very few mycelial or endospore elements tion rates as compared to two other gen­ were detected. era (Fusarium and Diplodia), considered to be seconda1y parasites in this particular discussion situation, could have been due to the fact that C. fimbriata was mainly present as The observed symptoms were very remi­ . chlamydospores in the ligneous tissues. niscent of those described on plane trees in USA and Europe, and on some fruit trees infested by Ceratocystis .fimbriata conclusion (plum, apricot and peach), especially in This is the first description of these quite California (DEVAY et al., 1968; MOLLER & unique symptoms in Citrus crops. C. fi m­ DEVAY, 1968; Mou.ER et al., 1969; BOSTOCK . briata seems to be the causal agent of & MIDDLETON, 1987) this new disease. The present results In plane trees, the fu ngus penetrates the should now be confirmed by pathogenic Photo 3 host through wounds in its aerial parts. analysis of C. .fimbriata in Citrus, e.g. Chlamydospore (alieuspore) C. .fimbriata mainly develops in the after experimental inoculation. Studies observed in Ceratocystis medulla1y rays of the infested tree (thus aimed at stalling the progress of this new culture derived from specimens explaining its flame-like appearance) and cut fro m areas around lesions. parasitic threat are under way to fu rther spreads to the pith. It then spreads in var­ determine some epidemiological compo­ ious directions and becomes deeply nents of the disease (etiology and propa­ implanted in the host. Tbe lesions spread gation). 9 rapidly, i.e. from 10 to 100 cm/month. The fu ngus is transmitted through disper­ sion of various kinds of debris from infec­ ted trees, e.g. sawdust from diseased trees references is highly infectious. Fungal inocula are BOSTOCK R.M., MIDDLETON G.E., 1987. most commonly transmitted on tools used Relationship of wound periderm formation to on different trees. Transmission by a few resistance to Ceratocystis fimbriata in potential insect vectors has also been almond bark. Phytopathology 77 (8): 1174- reported in USA (CRONE & BACHELDER, 1180. 1961), but this mode seems to be rela­ CRONE L.J., BACHELDER S., 1961. tively limited. Insect transmission of canker stain fungus, In stone fruit trees, C. .fimbriata is report­ Ceratocystis fimbriata F. platani. Phytopatho­ edly transmitted from the soil to pruning logy 51: 576 (Abs.). wounds or other injuries by insects DAVIS S.H., PETERSON J.L., 1973. (MOLLER & DEVAY, 1968). A tree wound dressing to prevent spread of In the Colombian situation, C. .fimbriata the Ceratocys tis causing canker stain Photo 4 seems to be maintained as chlamydos­ disease of the plane tree. Plant Dis. Rep. 57: C. fimbriata chlamydospores pores in the tree tissues, particularly in 28-30. observed in the xylem vessels the A' Ylem vessels (Photo 4). Identical of tree tissues. DEVAY J.E., LUKEZIC F.L., ENGLISH H., TRUJILLO E.E., MOLLER W.J., 1968. Ceratocystis canker of deciduous fruit trees. Phytopathology 58: 949-954.

MOLLER W.J., DEVAY J.E., 1968. Insect transmission of Ceratocystis fimbriata in deciduous fruit orchards. Phytopathology 58: 1499-1508.

MOLLER W.J., DEVAY J.E., BACKMAN P.A., 1969. Effect of some ecological factors on Ceratocystis canker in stone fruits. Phyto­ pathology 59: 938-942.

416 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards Research and Control Programmes Against Fruit Flies in Reunion

••••••••••••••••••••••••••• •••••••••• ••••••

introduction s1t1on sites and on the reproductive potential of this species. In recent years, the CIRAD-FLHOR ento­ mology laborato1y in Reunion has refo­ The importance of some visual stimuli, cused its activities on fruit flies, a group e.g. host colour and size, in determining of pests that have a major economic female oviposition site choices, was clari­ impact. The Natal fly, Ceratitis fied in the initial studies. Artificial wax (Pterandru.s) rosa Karsch, causes the most hemispheric domes (BOLLER, 1968; fruit crop damage on the island by far; PROKOPY & BOLLER, 1971) of given size this could be explained by the fact that it and colour were presented to females in is highly polyphagous and widely distrib­ choice experiments. Wild larval fruit flies uted. of the same origin were collected from infested guavas. The female fruit flies Three types of activity on this topic have used in most of the tests were naive with been under way since 1991: - baseline research, aimed at further char­ no prior contact with an oviposition sub­ acterizing the bioecology of this pest, strate. - applied research, aimed at improving The results showed a significant effect of supervised control techniques against fruit host size on the extent of ovipositions in flies, C. rosa, i.e. there were more eggs laid in - development activities, aimed at fully large-sized domes. Colour was also found informing fruit producers as to the situa­ to influence oviposition site choices; tion and distributing control compounds females generally preferred yellow and and equipment. red, and black to a lesser extent. Of all In 1991, a new Tephritidae species, seven colours tested, the highest numbers Bactrocera (Bactrocera) zonata Saunders, of eggs were laid in yellow domes. The was identified on the island. A control results also indicated no significant effect programme focusing on this pest was of background colour (white or black) in immediately set up in conjunction with the females' host colour choices. several other organizations. Moreover, it was found that precondition­ ing females on a red substrate had little effect on their ultimate substrate colour baseline research preferences. These studies are specifically focused on The studies also highlighted an important defining the underlying mechanisms that effect of olfacto1y stimuli in C. rosa ovipo­ guide C. rosafemales in choosing ovipo- sition site choices. In these trials, fruit

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 417 • e QUILICI

extracts were placed under the wax romonal marking generally begins late in domes out of the fruit flies' reach. In the afternoon, around 1700 hours. The choice experiments with odourless domes lek phenomenon that has already been and those smelling of peach (pulp) or reported in various Tephricliclae species, orange (peel), almost all eggs were laid in was observed for the first time in C. rosa. the latter domes. It seemed that other fac­ tors such as dome colour and size did not An ecological study on Tephritidae host limit the high attractivity of the orange­ plants in Reunion, aimed at updating data oclourecl domes, indicating that olfacto1y obtained about 20 years ago by ETIENNE stimuli are more important than visual (1982), has also been under way for stimuli. 2 years. Samples are being collected from a wide range of fruit hosts at many loca­ r The female f uit flieswere also very attrac­ tions and at different times of the year. r ted to other f uits such as rose apples, This survey revealed new host plants and Sizygium jamhos (L.) Alston. Bilberries dominant fruit fly species on the whole (Solanum auriculatum Ait.) were not as range of host plants. attractive, and the odour of guava 1 (Psidium cattleianum Sabine) was not at In addition, there has been an IOBC task all attractive to females. Otherwise, grape, force study on the responses of local which is not infested by C. rosa in field Ceratitis (Ceratitis) capitata Wiedemann conditions, induced some response under strains to pheromone marking (BOLLER et experimental conditions; damaged grapes al., 1994). were more attractive than sound fruit. Biological control studies were first focu­ In other tests, females did not respond, at sed on optimizing methods for mass rear­ the tested concentrations, to different ing local Tephritidae strains. In prepara­ chemical compounds used separately tion for acclimatization experiments, (acetic acid, limonene, linalool and fluo­ bioethological studies have been con­ rene). Laborato1y studies analysing ducted with an ovo-pupal parasitoicl, changes in volatile discharges from citrus Biosteres arisanus Sonan (Hym.: fruits (mandarin) during ripening (i.e. fruit Braconiclae), that was imported from odours at diffe rent maturity stages) rela­ Hawaii in early 1993 (generously donated 2 tive to female fruit fly responses are cur­ by Dr. E. Harris, USDA ). rently under way. They are aimed at fully defining the relationships between C. rosa and various host plants. applied and development­ Studies have also been carried out on oriented research other biological and behavioural aspects of C. rosa. The effects of temperature on Various tests have been carried out in the reproductive potential of the species recent years to improve supervised con­ has thus been investigated. For pre-adult trol techniques. The first were focused on stages, the thermal constant and zero­ perfecting sexual trapping techniques to point of egg and pupa development have monitor fruit fly populations. The effica­ been determined, i.e. 32.8 clegree-clays- cies of various types of traps, slow-release 11.l'C and 141.5 clegree-clays-12.6"C, sexual attractant dispensers, and diffe rent respectively. Other studies revealed that sizes of clichlo1vos insecticide strips female fecundity was maximal at 2o·c. (DDVP) were thus compared. Natal fly behaviour under semi-natural The results of several orchard tests carried conditions (large cage) was also studied. out since 1991 have confirmed the effi­ Some prelimina1y data were thus obtai­ ciency of the spot treatment technique (1) IOBC: International ned on the fly's activities and positions on (protein + malathion hydrolysate) against Organization for Biological the plant during the clay. Studies on the local fruit fly species present in Citrus Control of Noxious Animals sexual behaviour of the pest, especially and mango orchards. To assess the limita­ and Plants. the pheromonal marking of males, revea­ tions of this method, experiments were (2) USDA: United States led similarities with the known behaviour undertaken with other more susceptible Department of Agriculture. of C. capitata. However, in C. rosa, phe- fruit species (loquat, strawbeny guava

418 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY e

and peach); the results so far are not ve1y destruction of susceptible host fruits was convincing. Moreover, several orchard carried out around the outbreak area. tests were carried out to determine opti­ Network captures decreased progressively mal hydrolysate and insecticide concen­ during the second half of 1991 and were trations. Mortality rates at different com­ nil by Februa1y 1992. Only three flies pound concentrations were estimated by were trapped in 1993 and no new hanging sticky trays in the foliage of trea­ upsurges were noted. At the end of 1993, ted trees. the species was considered to have been completely eradicated from Reunion. A parallel programme was launched in Nevertheless, the fruit fly watch network 1991 to extend the use of supervised fruit must be kept up to enable ve1y quick fly control to all potentially applicable sit­ interventions if any more of these pests uations. The advantages of this type of are detected. pest control over the techniques currently used by fruit producers are: fewer treat­ ments (with the use of sexual trapping), lower control costs (reduced insecticide conclusion volumes, quantities and labour), thus Studies carried out over the last few years reducing harmful effects on non-target in Reunion by the CIRAD-FLHOR ento­ fauna. mology laborato1y have aimed at fully A widespread public awareness campaign characterizing the bioecology of (leaflets, posters, radio and TV ads and Tephritidae fruit crop pests on the island, videos), conducted in collaboration with particularly the Natal fly. Insect-plant rela­ the chamber of commerce of Reunion, tionships, pest behaviour and their natu­ was able to reach many farmers and ral enemies are priorities for future much of the general public. Essential research. equipment and products (traps, attrac­ Supervised control techniques are cur­ tants, etc.), which were partly funded by rently being extended in Reunion. They local governmental agencies (i.e. regional will eventually be combined with biologi­ and county councils), were distributed in cal control, which could be ve1y useful conjunction with many agricultural coop­ for treatment of areas with communities eratives. In 1992, sexual trapping was of wild host plants, and with biotechnical used to control about 450 ha of suscepti­ control, which will require further experi­ ble crops, and fruit flies were controlled ments. • on about 115 ha with spot treatments. A fruit fly pest watch network, involving sexual trapping with methyleugenol, was set up in Reunion by the mid-1990s. It aims at early detection of infiltrations by Bactrocera (Bactrocera) zonata Sanders, an Indian species that has been infesting fruit crops in neighbouring Mauritius over acknowledgements the past few years. A few flies of this spe­ cies were actually detected near the air­ This paper provides a brief summary of port in February 1991. A control program the research fo cuses of the CIRAD-FLHOR was immediately set up jointly with vari­ entomology la boratory team in Reunion. ous collaborating organizations (Service We would like to express our sincere de la protection des vegetaux, Federation gratitude to the technicians A. Franck, des groupements de defense contre les R. Manikom and C. Simiand, and to ennemis des cultures, Chambre d'agricul­ the fo llowing trainees who worked in ture de l"ile de la Reunion) to t1y to eradi­ our laboratory/ram 1991to 1993: cate this new pest. The sexual trapping K. Bonacina, A. Bonhomme, network was substantially reinforced in V. Bunge-Vivier, E. Ha ug, P. Labarriere, the zone and extended over the whole J.M. Ma rtin-Teissere, C. No rosomahefa, island. Simultaneously, an intensive cam­ A. Pepp uy, A. Pierru, L. Rivry, G. Rosso/in paign involving chemical control and and V. Sourdri.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 419 •• e QUILICI

references

BOLLER E., 1968. QUILICI S., 1989. An artificial oviposition device for the Euro­ Amenagement de la lutte chimique centre les pean cherry fly, Rhagoletis cerasi. J. Econ. mouches des fruits a la Reunion. In: Fruit Ent. 61: 850-852. Flies of Economic Importance, Proc. CEC / IOBC Intern. Symp., Rome, Italy, 7-10 April BOLLER E.F., HIPPE C., 1987. Rotterdam, Netherlands, R. CAVAL­ PROKOPY R.J., ENKERLIN W., KATSOYANNOS B.I., LORO ed., A.A. Balkema, p. 515-524. MORGANTE J.S., QUILICI S., DE CRESPO DE STILINOVIC, ZAPATER M., 1994. QUILICI S., TRAHAIS B., 1993. Response of wild and laboratory reared Improving fruit fly trapping systems in Ceratitis capitata Wied. (Dipt., Tephritidae) Reunion Island. In: Fruit Flies: Biology and flies from different geographic origins to a Management, New-York, USA, M. ALUJA et standard host-marking pheromone solution. P. LIEDO eds, Springer-Verlag, p. 235-242. J. Appl. Ent. 118: 84-91. QUILICI S., RIVRY L., ROSSOLIN G., 1994. ETIENNE J., 1982. Visual stimuli influencing the choice of ovipo­ Etude systematique, faunistique et ecolo­ sition site in Ceratitis rosa Karsch (Diptera: gique des Tephritides a la Reunion. PhD the­ Tephritidae). In: Current Research on Tropical sis, Paris, France, Ee. Pratique Hautes Fruit Flies and their Management, Proc. Etudes, 100 p. Symp. on Tropical Fruit Flies, 18-20 May 1992, Kuala Lumpur, Malaysia, H.S. YONG PROKOPY R.J., BOLLER E.F., 1971. and S.G. KHOO eds, p. 9-21. Artificial egging system for European cherry fly. J. Econ. Ent. 63: 1413-1417. • •••

420 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards The Fruit Fly Research Programme in New Caledonia

J.M. LEMONTEY •••••••••••••••••••••••••••••••••••••••••••••• F. MADEMBA-SY An important fruit fly research programme was set up in New Caledonia CIRAD-FLHOR Station de recherches fruitieres fo llowing the recent ban on ethylene dibromide treatments. It is aimed de Pocquereux at revitalizing the fruit and vegetable export sector. A progress report BP 32 98880 La Foa is presented. New Caledonia •••••••••••••••••••••••••••••••••••••••••••••• Fruits, vol. 49, n°5-6 p. 421-427 (English) introduction Te phritidae inventory p. 496-499 (French) A fruit fly research programme has been in New Caledonia set up in New Caledonia (Map p. 362) to Eleven fruit fly species have been identi­ address recent quarantine restrictions on fied in New Caledonia, including seven the export of local fruits and vegetables, endemic species (COCHEREAU, 1970; DREW, especially to countries such as New 1989; WHITE & ELSON-HARRIS, 1992). All of Zealand which are free of pest fruit fly these species, their known fruit hosts, species (Diptera: Tephritidae). Exports susceptibilities to different sex attractants, from New Caledonia to New Zealand and South Pacific distributions are sum­ were authorized until 31 December 1993 marized in Table 1. Two complementary on condition that the produce was pre­ strategies have been developed to certify treated with ethylene dibromide (EDB). the detection of all fruit flies and collect Fruit flies can be efficiently eliminated at further information on the full range of different development stages (e.g. eggs, plant hosts, these are: a sexual trapping larva) by treatment with this pesticide. network and systematic collection of sus­ However, New Zealand legislation has ceptible fruits and vegetables. now set the permissible EDB residue level at 0.1 ppm, thus representing an sexual trapping network overall ban on the use of this product. The sexual trapping network was set up Only fruits and vegetables with a con­ in 1990 and gradually extended through­ firmed non-host status for different fruit out New Caledonia; there are now 118 fly species, or those that have undergone traps at 41 different sites. Each Lynfield an authorized alternative treatment (heat, trap comprises a plastic container (1 I) cold, etc.), can currently be exported to with four holes, cotton soaked with a liq­ New Zealand. The French Territory of uid sex attractant and a dichlorvos insecti­ New Caledonia has assigned the CIRAD­ cide strip. The traps are attached at FLHOR fruit research station at Pocque­ human height under the tree foliage. reux the task of solving this produce Three sex attractants are used (Cue-lure, export problem. Methyl-eugenol and Trimedlure); the first two are known to effectively attract some A 4-year research programme 0993-1996) fruit fly species present in New has been set up under mixed financing Caledonia, and Trimedlure is used to (CIRAD/Territory of New Caledonia). The detect accidental introductions of Mediter­ present paper summarizes this pro­ ranean fruit flies ( Ceratitis capitata gramme, with emphasis on the first two Wiedemann). well-established phases, i.e. the sexual trapping network and Bactrocera tryoni Although fruit fly captures are not always (Froggatt) breeding project. representative of the actual field situation,

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 421 • • LEMONTEY and MADEMBA-SY

Table 1 Inventoiy of fruit flies (Tephritidae) in New Zealand (fromWHITE & ELSON-HARRIS, 1992; DREW, 1989 and DREW, pers. comm.).

Species Known host fruit Attractant Distribution

Bactrocera tryoni ve1y many Cue-lure Australia, N.C. (Froggatt) Polynesia Bactrocerapsidii citrns frnits, guavas, Cue-lure endemic N.C. (Froggatt) peaches, mangoes, ere. Bactrocera cu,vipennis citrns fruits, guavas, Cue-lure Vanuatu, N.C. (Froggatt) peaches, mangoes, etc. Bactrocera mucronis (Drew) unknown Cue-lure endemic N.C. Bactrocera wnbrosa breadfrnit, jackfrnit Methyl-eugenol Vanuatu, N.C., (Fabricius) Micronesia, East Asia Bactrocera ebenea (Drew) unknown Methyl-eugenol endemic N.C. Bactrocera aneuvittata (Drew) unknown unknown endemic N.C. Bactrocera perpusil/a (Drew) unknown Cue-lure (?) endemic N.C. Ba,etrocera.fu.lvifacies (Perkins) unknown unknown endemic N.C. Bactrocera. caledoniensis unknown Cue-lure endemic N.C. Ba.ctrocera sp.* Diospyros Ja .sciculosa. unknown endemic N.C. (?) (Ebenaceae) Bactrocera. near xa.ntbodes• unknown Methyl-eugenol Vanuatu (?), N.C. Dirioxa pornia(Walker) citrns fruits, peaches, unknown Australia, N.C. (secondary pest) ' identification in progress.

the sexual trapping network provides which are still relatively free of this spe­ information on the distributions and rela­ cies. The B. tiyoni risk potential should tive population densities of different fruit be stressed in the light of its rapid spread fly species (Fig. 2). B. t1yoni (Froggatt) subsequent to an accidental introduction and B. psidii (Froggatt) are often trapped from Australia in 1970 (COCHEREAU, 1970). in high numbers, i.e. 55% and 39% The bimonthly collections also provide respectively; note that these trapping fig­ data on fruit fly population dynamics. ures are means for all sites, but there are Trapping results at the Pocquereux station marked between-site variations. B. tryoni are summarized in Figure 3, highlighting seems to be spreading rapidly as it was the clear correlations between population captured in 98% of cases in Noumea; nev­ peaks and summer heat spells. The ertheless, only a few have been trapped curves are similar for each region, except on the Loyalty Islands (Mare, Lifou) for the urban zone around Noumea where fruit fly populations remain high B. u.mbrosa.3% others 3% throughout the year. This could be explained by the island-like location of Noumea and the fact that climatic varia­ tions are not as extreme as elsewhere in New Caledonia. All 11 characterized fruit fly species are captured regularly in the traps, even though there are veiy few B. t1yoni 55% specimens of some species, i.e. B. Ju lvffa­ cies (Perkins) and Dirioxa pornia (Walker). A few flies of a species close to Figure 2 B. xanthodes (Broun) have been trapped Percentages of fruit flies in the network set up on Mare Island. It is trapped in New Caledonia now being fully identified with the assis­ (mean of 108 traps). tance of the entomology laboratoiy of the

422 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY e

Queensland Department of Prima1y Industries (Dr. R.A.I. Drew). 40 35

fruit and vegetable collection 30 I» 0 The trapping network provides a rela­ tively easy means for drawing up a partial 25 .§ fruit fly inventory. However, the list obtai­ ned should not be considered exhaustive 20 2 since species that do not respond to the different sex attractants used are not 15 detected. This shortcoming was high­ p' lighted by the fact that some specimens 10 I» 0 of a seemingly new Bactrocera species (Drew, pers. comm.) were collected sep­ 5 arately on Diospyros fasciculosa (Ebena­ v ceae) fruits, but they have never been 0 ------·· 0 found in the traps. 28 32 36 40 42 46 52 4 811 12 •. 16______20 24 28 July December February May Wild and cultivated fruits and vegetables N° thus have to be collected at all sites weeks throughout the year. They are placed in

"hatching boxes" on a wet bed of saw­ Figure 3 dust. The sawdust is sifted regularly to B. tryoni, B. psidii, Variations in fruit fly trapping collect fruit fly pupa and adults, which B. curvipennis rates at the Pocquereux are then identified. research station (July 1993- and B. umbrosa rearing June 1994; mean of 3 traps). This campaign to detect fruit fly species that are not attracted by the sexual traps A rearing programme for the above-men­ is also useful for determining diffe rent tioned pest species was considered neces­ species of host plants. The fruit hosts of sa1y to produce enough flies for year­ six fru it fly species are still completely round reproducible and fully-controlled unknown. experiments. This list is being drawn up in collabora­ Fruit fly colonies were created from infes­ tion with the New Zealand Ministry of ted peaches (Prunus persica) for B. tryoni Agriculture and Fisheries (MAP). Econo­ and B. curuipennis, guavas (Psidium gua­ mically important fruit fly species, i.e. jaua) for P. psidii and jackfruit (Arto­ those that infest exportable fruits and veg­ carpus heterophyllus) for B. umbrosa. etables, are thus being identified. In prac­ tice, MAP considers that all fruit species B. tryoni rearing conditions belonging to the same plant family as a About 2 OOO adults/cage were kept in a known fruit host could be infested and temperature-controlled room (25"C ± 1 "C) thus should be tested. For instance, at about 70% relative humidity, under nat­ B. umbrosa (Fabricius) has been observed ural lighting with supplementary strong on Momordica sp., which means that it artificial lighting. The flies had a regular should be tested on all other Cucur­ supply of water, sugar, yeast hydrolysate bitaceae species present (e.g. zucchini, and the bacterial strain Klebsiella oxytoca. squash). This latter enterobacterium is commonly The list is needed as a basis for discus­ used in rearing Te phritidae fruit flies (LLOYD, sions on the exclusion of some species pers. comm.); it provides a protein from future host status studies. B. tryoni supplement which promotes the produc­ (Froggatt), B. psidii (Froggatt), B. curui­ tion of high quantities of viable eggs and LLOYD, pennis (Froggatt) and B. umbrosa sexual maturation (DREW & 1989). (Fabricius) are so far the only species Eggs are collected on artificial oviposition chosen; a programme has thus been set domes made of perforated plastic cylin­ up to rear these pests. ders (photographic film capsules), coated

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 423 • • LEMONTEY and MADEMBA-SY

inside with a larval medium containing medium used for B. umbrosa, and the fr esh banana (88.6%), Torula yeast extract lighting conditions in the rearing rooms. (11.1%) and nipagine (0.25%), an antimi­ B. curuipennis and B. umbrosa rearing crobial agent that effectively stimulates conditions are gradually being perfected egg laying in female fruit flies. The eggs and functionalized. So far ve1y few are collected in water and deposited on B. psidii flies have been successfully this larval medium. Larval pupation reared in the laborato1y. Nevertheless, occurs in wet sawdust spread under the mass numbers of these fruit flies should larval containers. The fruit fly rearing pro­ be collected from infested fruit during the cedures were extensively described in a next hot season; further rearing trials manual (CLARE and LEMONTEY, 1994). could then be undertaken.

B. psidii, B. curvipennis B. Tryoni life cycle and B. umbrosa rearing Fruit fly life cycles have to be fully under­ Conditions for rearing these three species stood in order to meet specific needs for are not as well established as for insects at different development stages. B. tryoni. There are differences with res­ Three thousand eggs that had been ovi­ Figure 4 pect to a number of factors, especially the posited in 2 h were thus deposited on Bactrocera tryoni life cycle 600 g larval medium. After hatching, at 25·c ± 1 ·c (on banana yeasts used to feed adult flies, the medium). potato/dehydrated carrot-based larval 100 larvae were randomly collected from the medium every 12 h for 2 weeks and their development stages were specified % (Pt, znd and 3rd instars). Changes in the 100 mean composition of the different larval stages on the nutrient medium are shown in Figure 4. Table 2 provides various bio­ 80 logical data on B. tryoni. Control of B. tryoni rearing and critical 60 information that has been obtained on its life cycle will facilitate further studies with this species. The same biological studies 40 will be conducting with other species once they can be efficiently reared.

20 host and non-host status o of plants 0 0.7 1.7 2.7 3.7 4.7 5.7 6.7 7.7 8.7 9.7 10.7 11.7 12.7 13 7 14.7 MAF drew up a list of different fruits and Days after oviposition vegetables to test on the basis of data col­ lected through the sexual trapping net­ -e- eggs + P' instar -e- znd instar ...... 3rd instar pupa -+ work and from collected wild and mar­ keted fruit (Table 3).

Table 2 Tests concerning B. t1yoni and B. curui­ Biological characteristics of B. tryoni (at 25°C ± 1 °C). pennis were staggered to coincide with the November 1994-February 1995 fruit­ N" eggs/female/24 h about 15 (15-30 days after oviposition) ing period so as to obtain high numbers Hatching rate about 70% of available gravid adult females. For the Hatching 41-53 h after ovipositioo two other species, the number of tests Pupation 8-14 days after oviposition carried out was dependent on the rearing Emergence of adults 21-29 days after oviposition success rates. Details on the studies con­ Sexual maturity about 15 days after emergence ducted for each fruit and vegetable with Length of the life cycle about 40 days each fruit fly species are given in Figure

424 •• Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY •

5; they met with the MAF specifications for the determination of fruit f1y host sta­ Table 3 tus (Ai'JONYMOUS, 1991). MAF list of fruits and vegetables for testing. Fruits and vegetables that were found to be non-hosts for all concerned fruit f1y Fruits and vegetables (varieties) Species to test species can be exported again without any type of postharvest treatment. lime (Tahiti SRA 58) B. fr., B. cur., B. ps., B. um. Alternative nonchemical treatments will litchi (local variety) B. 11"., B. cur., B. ps. be required for those that host fruit f1ies, mango (Kensington) B. um. e.g. mango was found to be a host for pineapple (Queen Tahiti) B. fr., B. cur., B. ps., B. um. B. tryoni, B. psidii and B. curuipennis. zucchini (Diamant) B. um. eggplant (Black beauty, Zebrina) B. tr., B. cur., B. ps. heat-treatment of mango squash (New Zealand variety) B. um. The efficiency of hot-air and vapour heat B. fr.: Bactrocera tryoni B. cur .. Bactrocem cu1vipennis treatments to eliminate various fruit f1y B. ps.: Bacfrocera psidii B. um.: Bacfrocem umbrosa. stages that infest fruit has been investi­ gated with several different fruit species. ARMSTRONG et al. (1989) thus demon­ been checked for effectiveness, but this strated the effect of high-temperature for­ will soon be done. Further analyses ced-air treatments of papayas infested should be carried out if any of the other with Ceratitis capitata (Wiedemann), three species shows higher resistance. Dacus cucurbitae (Coquillett) and D. dor­ There are several phases to these analy­ salis (Hendel) in Hawaii. Similar disinfes­ ses: tation studies with Bactrocera xanthodes - confirming lethal hot water bath treat­ (Broun) and B. melanotus (Coquillett) ment temperatures and times with artifi­ were carried out in the Cook Islands and cially infested fruits; led to a lifting of the ban on papaya - checking that the temperatures used are exports to New Zealand (WADDELL et al., not phytotoxic to the treated fruit; 1992 and 1993). In mangoes (cv - checking treatment efficiency in large­ Kensington), HEARD et al. (1992) demon­ scale tests. This involves exposing high strated the efficiency of vapour heat treat­ numbers of insects (30 OOO) to heat to cer­ ment of B. tryoni in Australia. In New tify that lethal temperatures established in Caledonia, disinfestation studies should previous tests actually cause 100% fruit f1y be carried out with B. tryoni, B. psiddi mortality. and B. curuipennis infesting cv Ken­ sington mango; B. umbrosa should also Our research is now focused on deter­ be assessed in this context if the host mining comparative resistances and the status of mango for this fruit f1y is found completed results for B. tryoni will be to be positive. published later. The first phase in the heat-treatment research involved comparing the resis­ tances of different fru it fly species at all conclusion development stages (i.e. egg, larva, l5l, In New Caledonia, postharvest treatments 2nd and 3rd instars) over an increasing are essential for export fruit and vegeta­ temperature gradient. This was carried ble crops because of the presence of fruit out by immersing eggs and larvae in hot flies in this country. Fruit f1y pest water baths for increasing periods of research carried out at the Pocquereux time, thus obtaining 0-100% mortality. station (New Caledonia) is fully in line MAF considers that the vapour heat treat­ with current world trends to develop ment developed in Australia could be alternative nonchemical postharvest treat­ used without modification, whereas our ment techniques. Scientific collaboration preliminary tests indicated high resistance with the Horticulture and Food Research of B. tryoni. The treatment has not yet Institute of New Zealand (Hort+Research,

Fruits, vol. 49 (5-6) / Special on Tropical Orchards : 425 • • • LEMONTEY and MADEMBA-SY

Laboratory test Adult cage

1 2 ------J t 24 h from oviposition [::] � � � Determination of the 500 g fruit to test 500 g known number x of females ' / host fruit required to obtain 500 g fruit to test 500 eggs in 24 h (ovipositions on domes) TEST CONTROL

oviposition, larval development no Non-host and adult emergence status

yes

Field test enclosed in a sleeve placed in test tree

24 h from oviposition

5 replications + control with known host fruit

Non-host no status oviposition, larval development Figure 5 and adult emergence Procedure for determining the yes fru it fly host status of various Host status fru its and vegetables.

ex-DSIR) will enable CIRAD-FLHOR to Pacific region. The know-how acquired benefit from their research experience in through the present research and the a similar disinfestation programme on credibility obtained by penetrating the papayas in the Cook Islands. renowned demanding market of New Official policies in New Caledonia and Zealand should facilitate future exports of economic imperatives have prompted various fruits and vegetables to other fruit and vegetable producers to export markets in the region, especially Japan. • their crops to other markets in the South

426 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY e

references DREW R.A.I., LLOYD A.C., 1989. Bacteria associated with Fruit Flies and their Host Plants. In: World Crop Pests: Fruit Flies - Their Biology, Natural Enemies and Control. Amsterdam, Netherlands, ROBINSON A.S. et Anonymous, 1991. HOOPER G. (eds), Elsevier Science NASS Standard 155.02.01.08: Specification Publishers, 3A, p. 131-140. for Determination of Fruit Fly Host Status as a Treatment. New Zealand Ministry of HEARD T.A, HEATER N.W., PETERSON P.M., Agriculture and Fisheries, 17 p. 1992. Relative Tolerance to Vapor Heat Treatment ARMSTRONG J.W., HANSEN J.D., HU B.J.K., of Eggs and Larvae of Bactrocera tryoni BROWN S.A., 1989. (Diptera: Tephritidae) in Mangoes. Journal of High-Temperature, Forced-Air Quarantine Treat­ Economic Entomology 85 (2): 461-463. ment for Papayas Infested with Tephritid Fruit Flies (Diptera: Tephritidae). Journal of Econo­ WADDELL B.C., CLARE G., MAINDONALD J.H., mic Entomology 82 (6): 1667-1674. 1992. Postharvest Disinfestation of Bactrocera CLARE G., LEMONTEY J.M., 1994. melanotus and B. xanthodes in the Cook Fruit fly Rearing Management Manual for Islands: Reports 2. HortResearch Client Bactrocera tryoni, B. curvipennis, B. psidii Report No.92/89. Auckland, New Zealand, and B. umbrosa. HortResearch Client Report 41 p. No. 94/65. Auckland, New Zealand, 36 p. WADDELL B.C., CLARE G., MAINDONALD J.H., COCHEREAU P., 1970. 1993. Les mouches des fruits et leurs parasites Postharvest Disinfestation of Bactrocera dans la zone indo-australo-pacifique et parti­ melanotus and B. xanthodes in the Cook culierement en Nouvelle-Caledonie. Cahiers Islands: Reports 3. HortResearch Client ORSTOM, serie Biologie, 12, 15-50. Report No.93/270. Auckland, New Zealand, 70 p. DREW R.A.I., 1989. The Tropical Fruit Flies (Diptera: Tephritidae: WHITE I.M., ELSON-HARRIS M.M., 1992. Dacinae) of the Australasian and Oceanian Fruit Flies of Economic Significance: Their Regions. In: Memoirs of the Queensland Identification and Bionomics. Oxon: C.A.B. Museum, Brisbane, 26, 521 p. International, 601 p. • •••

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 427 Inventory of Insect Fauna Specific to Cultivated Fruit Trees of Northern Cote d'Ivoire

K. N'GUETIA •••••••••••••••••••••••••••••••••••••••••••••• DFA, Station de Lataha The highly diversified insect fa una colonizing five types of fruit trees BP 856 in northern Cote d'Ivoire was inventoried. Korhogo Cote d'Ivoire ••••••••••••••••••••••••••••••••••••••••••••••

Fruits, vol. 49, n°5-6 p. 428-429 (English) p. 500-501 (French) aims These insects were first identified from available documents or collections and The present study was aimed at: then confirmed or characterized by a spe­ - identifying all species of insects living cialist. on cultivated fruit trees in northern Cote The insects were prepared and classified d'Ivoire, in a collection. - determining the distributions of each species on the main fruit trees cropped in species distributions on different the region. host fruit trees Host fruit trees were noted for each insect species at each sampling. A table was materials and methods drawn up with each insect species and their hosts. insect collection Insects were collected with sweep nets in results experimental plots at the Lataha research station (Korhogo region, Cote d'Ivoire). Table 1 provides a list of insect species They were sampled from five different identified during the present survey and r types of host fruit trees: mango, citrus, the f uit trees from which they were col­ guava, papaya and cashew. lected. The insect species were either collected The insects were collected in the foliage, r between rows of plants in nurseries and on one specific type of f uit tree (Table 2), between rows of adult trees in orchards. or on two, three or four of the five types of trees. Collected insects were killed with acetic ether and identified by the naked eye or Very few insect species were specific to under a dissecting microscope. only one type of fruit tree. They were generally specific to one or two trees; heteropterans and coleopterans were criteria assessed exceptions, with more than two species species identification from these orders collected on more than one type of the fruit trees studied. The For each sampling, collected insects were distributions were as follows: given a code number corresponding to - six heteropterans and three coleopte­ various morphological features. They rans collected on both mango and citrus, were then temporarily referred to as "spe­ - three coleopterans on mango and cies". papaya,

428 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY •

- three coleopterans on citrus and guava, - four heteropterans on papaya and Table 1 cashew, Insect species collected on five different types of fmit trees in northern - eight heteropterans on mango, citrus Cote d'Ivoire. The results are pooled on an insect order basis. and guava, ° - six heteropterans on mango, citrus, Mango Citrus Guava Papaya Cashew N different guava and papaya. species The distributions of insect species col­ Heteroptera 50 49 31 32 25 96 lected on the five fruit trees studied were Coleoptera 26 17 15 10 10 53 as follows: Diptera 18 12 8 9 14 44 - four heteropterans: Neza ra viridula Hymenoptera 16 20 10 7 7 51 (Pentatomidae), Agonoscelis versicolor Orthopt era 1 1 2 2 3 (Pentatomidae), Aspavia annigera Lepidoptera 5 4 3 0 1 11 (Pentatomidae) and a Coreidae species, Homoptera 8 2 2 1 - one coleopteran: Lagria villosa (Lagrii­ 4 3 3 dae), Odonata 0 1 4 0 0 5 - one dipteran: He nnetia sp. (Stratio­ Neu'roptera 2 3 0 0 0 3 myidae), Dictyoptera 0 1 0 0 2 - one orthopteran: Zonocerus variegatus (Acrididae). One-third of the insect fauna observed on Table 2 cultivated fru it trees in northern Cote Insects specific to different types of cultivated fruit trees of northern d'Ivoire were heteropterans, next came Cote d'Ivoire. coleopterans, dipterans and hymenopte­ rans. Species of these orders were found Mango Citrus Guava Papaya Cashew on all of the fruit trees studied, but see­ med to prefer mango and citrus. The last Heteroptera 10 16 6 8 5 quarter of the insect fauna was repre­ Coleoptera 13 6 5 6 5 sented by six other species that seemed Diptera 14 9 5 3 8 highly specific to certain fruit trees. Hymenoptera 13 14 7 5 5 Species collected in this survey are not all Orthoptera 1 pests; some are predators of other insects, Lepidoptera 4 3 2 for instance: Homoptera 5 1 1 3 - Reduviidae and Lygaeidae (Geocoris sp.) Odonata 1 4 bugs, Neuroptera 1 - Coccinellidae (Exochomus sp., Cydonia Dictyoptera 1 vicina) coleopterans, - Asilidae (Promachus sp.) and Syrphidae (Xanthogrammaaegy ptium) dipterans, - various hymenopteran species, rans (dominant group), coleopterans, - Libellulidae odonatan species, Ch1yso- dipterans and hymenopterans. Low num­ pidae neuropteran species, Mantidae dic­ bers of orthopteran, lepidopteran, tyopteran species, etc. homopteran, odonatan, neuropteran and Other insect species lick or suck nectar or dictyopteran species were also detected. other plant exudates. These are generally The insect species are generally pests, dipterans (Bombyliidae) or various hyme­ predators or pollinators. nopterans involved in flower pollination. Further studies are to be carried out to supplement the present preliminary results; they will focus on: conclusion - identifying further species, This preliminary study highlighted the - assessing insect population densities on insect fauna that colonize the main culti­ different host plants, vated fruit trees in northern Cote d'Ivoire. - evaluating the economic impact of vari­ This fauna mainly comprises heteropte- ous insects on fruit crops. e

Fruits, vol. 49 (5-6) / Special on Tropical Orchards •• 429 Inventory of Insect Fruit Pests in Northern Cote d'Ivoire

K. N'GuETTA •••••••••••••••••••••••••••••••••••••••••••••• DFA, Station de Lataha Many insects that develop in fruits were inventoried and identified fo llowing BP 856 preliminary studies that were carried out in northern Cote d'Ivoire. Korhogo Cote d'Ivoire ••••••••••••••••••••••••••••••••••••••••••••••

Fruits, vol. 49, n°5-6 p. 430-431 (English) p. 502-503 (French)

aims criteria assessed The present study had three main objec­ species identification tives: The insect species were identified as des­ - to inventory all insect fauna of the cribed by N'GUETTA (1995). region that develop in fruit, - to identify their specific natural ene­ species quantification mies, The total number of insects obtained at - to determine population sizes for each each sampling was noted for each species species on different host plants. and host plant.

materials and methods results insect collection species distributions Twenty-seven insect species were col­ The insects were collected on infested lected on fruit, most were dipterans, lepi­ fruit harvested in experimental plots at dopterans and hymenopterans. the Lataha research station (Korhogo region, Cote d'Ivoire). They were col­ Of 16 dipteran species collected: lected from five different types of fruit: • eleven belonged to the Tephritidae family: mango, citrus, guava, papaya and cashew - Ceratitis species: C. (Pterandrus) ano­ apples. nae, C. (Ceratalaspis) cosyra (Walker), C. (Pardalaspis) punctata (Wied), insect rearing Ce ratitis spp., - one Dacus sp., Ripening orchard fruits were placed in • one belonged to the Stratyomiidae fam­ the front compartment of larval rearing ily: Hermetia sp., cages until completely decomposed. Eggs • four have not yet been identified. and larvae in the fruit continued their development. Of four lepidopteran species collected: • one belonged to the Oletheutidae fam­ Fully developed larvae left the fruit and ily: Cryptophlebia leucotreta, pupated in a layer of sand in the second • the three others (two collected on citrus compartment of the cage. The sand was and one on guava) have not yet been sifted to collect nymphs and other pread­ identified. ult insects. These developing insects were put in emergence cages and reared to The six hymenopteran (Microhymenop­ adulthood. tera) species collected have not yet been

430 • Fruits, vol. 49 (5-6) / Special on Tropical Orchards • CROP PROTECTION / ENTOMOLOGY e

identified. They are probably parasitoids also detected in high quantity. The larvae and/or hyperparasitoids. seemed to feed by scavenging since they were found in decomposing fruit waste. The fruit-host distribution of the However, the exact fruit pest roles of 11 Tephritidae species collected was as these species have not yet been deter­ follows: mined. • six species on mango only, four of these were Ceratitisspecies, Two Tephritidae species were found to • one species was found on mango and be the main guava pests, i.e. Ceratitis guava: C. capitata, cosyra and C. annonae. C. capitata, ano­ • one species on guava and papaya: ther species of this family, was only Dacus sp., detected in the last cv Beaumont guava • two species on mango, guava and harvest; its role should now be deter­ papaya: C. (Pterandrus) annonae and an mined. unidentified species, • one species was found on all of the fruit hosts: C. (Ceratalaspis) cosyra. conclusion One lepidopteran species, Cryptophlebia In the present study, the types of insects leucotreta (Oletheutidae), was collected that develop in fruit harvested in northern on citrus and guava. The other still Cote d'Ivoire were determined, along unidentified species seemed to be specific with their preferred fruit hosts. Further to fruit hosts. studies are under way to supplement The microhymenopteran species were and/or confirm the present results, they specific to mango, guava and papaya; focus on: only one species was found on both - identifying and studying collected insect mango and papaya. species, - assessing the sizes of all insect popula­ tions collected, - analysing the population dynamics of Insect quantities/species each species, /host plant - defining various storage techniques, - identifying parasitoids and their hosts. Only dipteran species on two fruit hosts e were considered in this analysis: papaya (cv Solo) and guava (cvs Beaumont and Huek-Now). On cv Solo papaya, Dacus sp. (Teph­ ritidae) was found in the highest quan­ references tities in all harvested fruit, it thus seemed N'GUETIA K., 1995. to be the most important and constant Inventory of insect fauna specific to papaya fruit pest. A still unidentified cultivated fruit trees of northern Cote Stratiomyidae species (Hermetia sp.) was d'Ivoire. Fruits, 49 (5-6): 428-429.

Fruits, vol. 49 (5-6) / Special on Tropical Orchards • 431