Study tour of South Africa, New Zealand, the UK and the USA (also incorporating Chile)

Transcription

1 Study tour of South Africa, New Zealand, the UK and the USA (also incorporating Chile) Milestone report for AV Rootstock Improvement for the Avocado Industry A Whiley and G Anderson Sunshine Horticultural Services Pty Ltd Project Number: AV01007

2 AV01007 This report is published by Horticulture Australia Ltd to pass on information concerning horticultural research and development undertaken for the avocado industry. The research contained in this report was funded by Horticulture Australia Ltd with the financial support of the avocado industry. All expressions of opinion are not to be regarded as expressing the opinion of Horticulture Australia Ltd or any authority of the Australian Government. The Company and the Australian Government accept no responsibility for any of the opinions or the accuracy of the information contained in this report and readers should rely upon their own enquiries in making decisions concerning their own interests. ISBN Published and distributed by: Horticultural Australia Ltd Level 1 50 Carrington Street Sydney NSW 2000 Telephone: (02) Fax: (02) horticulture@horticulture.com.au Copyright 2002 Horticulture Australia

3 Study Tour of South Africa, New Zealand, the UK and the USA (also incorporating Chile) A.W. Whiley and 6. Anderson *'W^ys^ Apr-May 2002 Horticulture Australia AV»CAD«

4 Contents Study Tour Highlights 2 Information Dissemination 3 Implications and Recommendations for the Australian Avocado Industry 3 Avocado Nurseries and Rootstock Cloning 4 Westfalia (South Africa) 4 Allesbeste Nursery (South Africa) 6 Pine Island Nursery (Florida) 10 ACW Nursery (California) 11 Greenleaf Nursery (California) 12 Huerto California Nursery (Chile) 13 Rootstock Status of Countries Visited 15 California 15 Rootstock breeding programme 15 Florida 16 New Zealand 16 South Africa 17 Rootstock breeding programme 18 Plant Growth Regulators 19 Apogee 19 Naphthalene-acetic acid 20 Canopy Management 21 South Africa 21 Florida 22 California 22 Ultra-high density orchards 22 Avocado Sunblotch Viroid 23 Packaging and Marketing Avocados 25 Packaging in South Africa 25 Preparing and Exporting 26 The RPC Trail in the UK 26 Chain Store Displays in California 30 Marketing in New Zealand 31 Contact/People Visited on the Study Tour 32 Appendices Acknowledgements This study tour was funded by Horticulture Australia Ltd and The Australian Avocado Grower's Federation. Dr Tony Whiley and Mr Graham Anderson visited South Africa, the USA and New Zealand. The Chilean Comite de Paltas funded a visit to Chile for Dr Whiley where additional information was gained.

5 2 Study Tour Highlights Highlights from the South African, UK, USA and New Zealand study tour include: 1. Visits were made to avocado nurseries where elite rootstocks were being commercially cloned. These included Westfalia Nursery in South Africa where 'Duke 7' and 'Merensky 2' were being cloned using the Brokaw "nurse seed" technique; Allesbeste Nursery in South Africa where the Brokaw "nurse seed" technique had undergone significant modification and micro-cloned rootstocks of 'Duke 7' and 'Velvick' were being produced; ACW Nursery in California which was using a modified technique similar to Allesbeste Nursery for the production of cloned 'Duke 7'; and Greenleaf Nursery in California which was also using a modified clonal technique similar to Allesbeste nursery for the production of 'Duke T clones. All nurseries visited indicated that they would like to participate in the exchange of new information on cloning and rootstocks as it develops. 2. Negotiations were held with Dr Stefan Kohne of Merensky Technological Services for access for ANVAS nurseries to the newly developed Phytophthora root rot tolerant rootstock Merensky 2. The rootstock is already in Australia under test agreement with ANFIC but Dr Kohne gave assurances that any commercialisation agreement in Australia would be inclusive of ANVAS nurseries (a written confirmation will be prepared by Dr Kohne). 3. A new product with fungicidal action has been developed and commercialised by Merensky Technological Services. The product called Westfalia Biocoat is a formulation of natural glycols and alcohols and is effective in controlling the postharvest diseases stem-end rot and anthracnose. The product can be safely used on fruit going into chemical-sensitive markets such as Europe. {Note: since returning to Australia QUI has arranged importation of this product for testing). 4. There is a growing demand for organic products in Europe and most of the farms visited in South Africa are targeting this market through managing some of their orchards to comply with European standards. Organic accreditation is only given following an on site inspection by the European certification body. 5. The germplasm block at South Coast Field Station, University of California was inspected and six Guatemalan types identified as potential rootstock material for Australian conditions. These will be sunblotch-viroid tested by the University of California prior to their introduction into Australia. 6. A visit was made to the USDA at Cutler Ridge, Miami where they are using AFLP technology to genotype avocado cultivars. There may be an opportunity for this group to apply their technology to avocado material in Australia for rootstock identification and this is currently being explored. 7. Our knowledge on ASBV has steadily increased as new, more sensitive detection methods have been developed. Centres were visited in Florida and California where PCR technology has been developed for ASBV detection. With current knowledge it is recommended at this point in time that a zero-tolerance policy should be adopted for all trees being propagated and sold from ANVAS nurseries as soon as practicable and this be enforced through the routine sampling of leaves for ASBV analysis from nursery trees. This policy should be supported by rigorous testing of nuclear and multiplication blocks and that the Australian industry should consult with the New Zealand industry to identify a reliable and affordable indexing service. 8. A visit was made to the Tissue Culture facility at the University of Florida, Homestead (Dr Richard Litz) where research is being carried out to "clean up" sunblotch viroid from

6 3 infected cultivars. We also learnt from this group that in tissue culture avocado cultivars are very protocol specific. This may have implications in micro-cloning as differences in the success rate between rootstocks have been reported from nurseries involved in cloning. Visited Mack Multiples in the UK who are a large importer of avocados from most major growing countries that supply Europe. They repack and supply all UK supermarkets using innovative pre-packages and nearly 100% returnable plastic containers (RPC's). Fruit is sprung and machine graded with a non-damaging firmness tester. They are also experimenting with using RPC's to import fruit directly in from the country of origin. Information Disseminated 1. This report will be supplied to members of the AAGF R&D committee and all members of ANVAS who will have particular interest in the nursery technology contained herein. 2. A presentation will be made to the AAGF R&D committee. 3. A series of articles will be written based on information in this report and published in the industry journal "Talking Avocados". 4. This report is in the public domain and can be accessed from Horticulture Ausytralia Ltd. Implications and Recommendations for the Australian Avocado Industry 1. The Australian Avocado Industry has fallen behind other countries in progressing genetic improvement in material planted in orchards. Gains have been made by California and South Africa with improved rootstock technology and the Australian industry needs to improve on the material that is currently being sold through ANVAS nurseries. In many cases there is no field-performance history on seedling lines being used. 2. Cloning technology should be thoroughly researched and all ANVAS nurseries made aware of the methodology as soon as reliable systems are available to clone "locally" selected roostocks. 3. The Australian industry should adopt a zero tolerance for avocado sunblotch viroid or its variants in any new material being sold through ANVAS nurseries. 4. Returnable Plastic Crates are being successfully used in South Africa and the UK and should be investigated for use by the Australian avocado industry. 5. Greater knowledge should be acquired of the California Avocado Commission's instore promotion programme.

7 4 West/alia Nursery (South Africa) Avocado Nurseries and Rootstock Cloning Avocado tree propagation has been formerly under control of the Westfalia Estates production unit but has recently been moved across to Merensky Technological Services headed by Dr Stefan Kohne. Merensky Technological Services have appointed Ms Erica Faber as their nursery manager and she brings considerable experience from both citrus tree and ornamental plant production to the job. Since entering into clonal rootstock propagation the nursery has produced about 1 million grafted trees on cloned rootstocks which have been planted on its own estates or sold to South African growers. More recently nursery trees have been exported to Kenya for use in the developing avocado industry of that country. In years of strong demand for nursery trees Westfalia produce about grafted trees on cloned rootstocks. Past demand has mainly been for 'Duke 7' but the nursery is currently changing across to 'Merensky 2' (a local selection) that has better Phytophthora root rot tolerance and improved productivity when compared with 'Duke 7'. 'Ettinger' is the main source of "nurse" seed material being available early in the autumn (Table 1). Seed is picked from the trees when mature and removed from the flesh while the fruit are still firm. The fruit is cut around the equator with a sharp knife and broken open (Fig. 1). Due to the stage of unripeness the seed coat comes cleanly away from the seed which is then hot water treated at 50 C for 30 minutes. Fig. 1 Removing 'Ettinger' seed from the fruit. Westfalia Nursery uses 'Ettinger' as a 'nurse' seed in the cloning process. Note the seed is removed when the fruit is firm and the seed coat comes away cleanly from the seed. On cooling the seed is dusted with a Benlate /Dithane mix (1:1) and then planted into folded nursery bags containing a composted pine bark media (Fig. 2). The Westfalia nursery is using the "Brokaw" method for producing cloned trees wherein the "nurse" seedling is retained as a composite part of the nursery tree right through to point of sale.

8 5 The cloned rootstock section is separated from the "nurse" seedling through the use of a "strangulation" washer placed at the base of the etiolated rootstock growth (Fig. 3). illlli i!!.. Fig. 2 Nursery bags folded in preparation for planting a 'nurse' seed. Once "nurse" seeds have been grafted to the elite rootstock they are placed in a dark room for etiolation. The room is held at approximately 27 C and well ventilated to prevent fungal growth on etiolating shoots. The etiolated growth is allowed to reach cm before removal from the room. The base of the Fig. 3 Strangulation washer placed at the base of the etiolated rootstock to give separation from the "nurse seed". shoot is treated with 0.75% KIBA, a strangulation ring fitted and the fold bag pulled up and potting media added. After about 100 mm of new growth has occurred on the cloned rootstock the fruiting scion is grafted to the tree. Following the hardening of new growth from the scion the young trees are transplanted into large (6-7 L) containers (Fig. 4) where they are grown through to the point of sale. The young transplanted trees are sprayed with Vapour Guard to reduce transpiration and placed under 60-80% shade for 2-3 weeks before shifting to a higher light regime. Fig. 4 Preparation of transplanting sites in nursery bags (left) in readiness for transplanting young, grafted cloned rootstocks for growing on the saleable size.

9 6 Allesbeste Nursery (South Africa) Allesbeste Nursery is located on the outskirts of Tzaneen in South Africa and is owned by Dr Andre Ernst who completed his PhD at the University of Pretoria studying avocado propagation. The nursery manager is Gerhard (Gerry) Erasmus. About 70% of the nursery trees produced by Allesbeste Nursery are on clonal 'Duke 7' rootstocks whilst the balance are on seedling rootstocks, mostly 'Zutano'. The nursery has strict hygiene standards throughout its operation (Fig. 5) with the nursery divided into zones. Disinfestation of the hands and feet of personnel moving between these zones is mandatory. Sodium hypochlorite (2.5%) is used daily to sterilise paths and work areas (very corrosive). Allesbeste Nursery uses composted pine bark as the main constituent of potting media's for the production of avocado nursery trees. The pine bark is supplemented with dolomitic lime to supply Ca and Mg and to raise the ph to around 6.0. Although the material is bought premixed with added macro and micro nutrients the nutritional content of the media is unreliable and is supplemented with additional fertiliser by Allesbeste Nurseries. Granular elemental sulphur (95% formulation; Sulfur 95 from Ocean Agriculture Pty Ltd at is added to the potting media at the rate of 1.5 kg m" 3. The composted pine bark is mixed with coarse river sand and soil at the ratio of 60/20/20. This media is used in Fig. 5 Main entrance to Allesbeste Nursery, Tzaneen; note covered copper fungicide footbath. the "nurse" seed tubes for the production of seedling trees used to graft to rootstocks for cloning and for the large planter bags where trees are grown through to saleable size. Potting media's are sterilized with methyl bromide at g m" J (Fig. 6). Fumigated soil is covered for 2 days and then aerated for about 2 hours before filling nursery bags. The seeds are planted in bags the next day without any phytotoxicity problems from methyl bromide. Allesbeste Nursery has found that the best results with rooting cloned rootstocks are achieved going into late summer and autumn (no reason for this was apparent). For clonal production the 'nurse' seeds are planted in February/March (as soon as available) with grafting commencing as soon as seedlings are large enough to take the rootstock scion wood. Grafting this union is carried out with PVC tape rather than using Parafilm as this is believed to stimulate rooting of the clone after Fig. 6 Potting media being prepared for sterilisation at Allesbeste Nursery, Tzaneen, South Africa. The soil in the concrete beds is covered with a plastic tarpaulin and fumigated with methyl bromide at 5.5 g m" 3 for two days.

10 7 etiolation (possibly due to additional constriction at the union as the stem diameter increase with growth of the plant). Shade cover is 60% on seedlings and 40% on the main production area. 'Zutano' and 'Ettinger' are used as "nurse" seed by Allesbeste Nurseries for propagation of clonal rootstocks (Table 1). Fruit for seed use is harvested from the tree but as a precautionary measure the surface of fruit is sterilised by treatment with 1.5% sodium hypochlorite when it is brought to the nursery. When the fruit has softened the seed is extracted, washed and then airdried for a couple of days to crack the seed coat. The seed is then hot water treated (30 min at 50 C), dipped in prochloraz and planted into black poly nursery sleeves. As temperatures get cooler going into autumn/winter, seed is placed in the heated etiolation room to assist germination but is moved to a greenhouse shortly after shoots appear (can graft etiolated growth if necessary). When "nurse" seedlings are large enough they are grafted with the elite rootstock selected for cloning (to date mostly 'Duke 7' for commercial production in South Africa) (Fig. 7). As buds on the rootstock scion begin to break plants are transferred to the etiolation room where they are held at 25 C until about 300 mm of growth has been made (Fig. 8). Plants are occasionally sprayed with copper oxychloride to manage fungal growth and insecticides to control mealy bugs and aphids. Nursery stock is managed on a computer database and annual batches of plants are always kept separate. The etiolation room used by Allesbeste is built with concrete blocks with an insulated ceiling but is also well ventilated. Fig. 8 Etiolation room at Allesbeste Nursery (left) with strong etiolated growth being produced from the Duke 7 rootstock material (right). Following etiolation the base of the shoot is sliced longitudinally with a sharp knife on one side of the stem removing a sliver of bark thereby exposing the cambium tissue. The cut region is then treated with 0.7% IB A in the first crop of shoots (0.5% IB A in the second crop of shoots) using one drop of solution from an eye-dropper (Fig. 9). The "nurse" seedling is kept growing through the application of a slow release fertilizer to the top of the sleeve and plants are watered daily.

11 8 Fig. 9 Cutting the base of the etiolated shoot in preparation for the IBA application. On removal from the etiolation room the plants are held under heavier shade for 3-6 weeks during which rooting usually occurs. The fruiting scion is grafted cm above soil level. The apical bud is removed from the scion when grafting both the candidate rootstock and fruiting scions to encourage branching. The cut surfaces are painted with Pancil- Plus, Sanachem Pty Ltd, PO Box 1454, Durban 4000, South Africa. The grafted, rooted clone is clipped from the nurse seed when 3-4 small leaves have been produced on the fruiting scion (Fig. 10). Clipping is carried out at this stage to reduce the transpiration shock following removal from the nurse seed. Once cut from the nurse seed the young, rooted cloned plant is grown on under greenhouse conditions. If force-grown it takes 8-12 weeks before planting out in the big pots. P.?.**.". * * ' - ^ 1.* «V < 1,&*:«^ Fig. 10 Micro-cloned Duke 7 rootstock grafted to 'Hass' at the stage plants are clipped from the "nurse" seedling (left); and a grafted micro-cloned rootstock showing root development in potting mix. Micro-cuttings can be managed in the small tubes for up to one year. When growth in microtubes is being forced plants are fertigated twice weekly with urea/kn03/nh4s04 taking care that the fertigation solution doesn't exceed 150 ms m" 1. Plants from the micro-tubes are transferred to 8 L nursery bags with extra holes cut in the bottom when the scion growth is fully expanded and hardened (Fig. 11). When transplanting the micro-tube tree a hole is made in the potting media with a coring tool large enough to allow the micro-tube plant to be slipped into place. After making the hole but prior to planting 10 g on slow release fertiliser is dropped in the hole and covered with potting media so as the roots of the micro-tube tree do not come into direct contact with the fertiliser.

12 9 Budwood supply for grafting is managed by cutting trees back to stimulate regrowth beginning in January. A group of trees are successively pruned back on a monthly basis so the mature wood with vegetative buds is available throughout the grafting season. Scion wood is not generally stored under refrigeration apart for immediate needs. 'Hass' trees grafted to cloned 'Duke 7' are usually sold for Rand (about $4.00 AUD) or 14 Rand (about $2.50 AUD) for trees on seedling rootstocks (Fig. 12). Fig.ll Transplanting micro-cloned 'Duke T rootstock grafted to a 'Hass' scion into 8 L nursery bags to grow through to a saleable size: left - making and fertilising the hole; middle - removing the micro-cloned from the nursery tube; right - transplanted tree. Fig. 12 'Hass' grafted to cloned 'Duke T rootstocks ready for sale (left and right) showing the root development in pots (centre).

13 10 Table 1 'Nurse' seed varieties used by nurseries producing cloned avocado rootstocks. Variety Nurseries ACW Allesbeste Brokaw C&M Greenleaf Huerto California Westfalia 'Ettinger' 'Lula' 'Reed' 'Zutano' Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Pine Island Nursery (Florida) Pine Island Nursery is one of the larger avocado nurseries in Florida servicing the commercial requirements of growers. The nursery is managed by Colleen Boggs and Eric Tietig. Avocado production in Florida is based on seedling rootstocks with some 25 different cultivars grown commercially. West Indian varieties do not store well on trees when mature so a longer season is obtained through choosing cultivars with different maturity times. Seeds of' Waldin' are used for propagation of all varieties. However, there is no comparative data on the use of seedling 'Waldin' as a rootstock with other seedling material. Seeds of 'Lula' have been used in the past but the seed size is small resulting in low-vigour nursery trees and it is difficult to graft to some varieties hence it is not popular with the nursery industry. 'Waldin' seed costs about 24 cents US each with a further 32 cents US to remove the seed from the fruit. No scarification is carried out although the seed coat generally is removed when extracting the seed from the fruit. The seed is planted directly following removal from the fruit and has a 95% strike rate. Avocado trees are grown in large containers (12 L) and are sold for $8-50 to $12-00 US depending on tree size when collected from the nursery. The main planting period is from August through to October (autumn). The nursery itself was basic with all of the production being carried out on weed mat laid directly on the ground (Fig. 13). Soil was not sterilised and granular pre-emergence weedicides were applied to the top of the pots for weed control. Nutrition to trees was applied via a granular slow-release fertiliser with an NPK ratio of Fig. 13 Avocado production at Pine Island Nursery, South Florida

14 11 ACWNursery (California) The ACW nursery is located near Fallbrook in southern California on the commercial avocado property of Dr Bill Arteberry but the nursery operation is supervised by Mr Reuben Hofshi com). The nursery was established to primarily service the requirements of the ultra-high density orchards being established on the property. 'Zutano' is most commonly used as the "nurse seed" and the largest seeds are selected and are not deliberately scarified following heat treatment, although seed coats generally come off during subsequent handling. At times 'Reed' is used as a "nurse seed". During propagation, the greenhouse is maintained at C and all stages of the cloning process are carried out on heated propagation tables that are held at about 27 C. Once reaching sufficient size, the "nurse" seedling is grafted with a scion of the selected rootstock and allowed to grow until a shoot begins to develop (approximately 14 days). The tree is then placed in darkness and the shoot etiolated until reaching about 400 mm in length (7-14 days) (Fig. 14). After the plant is removed from the dark, the stem is vertically scraped completely around for approximately 100 mm along the etiolated section and the area treated with a 1% IBA/0.5% NAA solution (Fig. 15). A sliced, clear plastic cup (240 ml) or clear plastic sleeve is taped around the stem with the scored area about one-quarter of the distance below the top of the cup/sleeve. The cup/sleeve is filled with sterile rooting medium and taped to a stake for support (Fig. 16a). The transparency of the container allows visual Fig. 14 Etiolation of the rootstock shoot inspection of root development so that progress can following grafting to the nurse easil be y monitored. Under optimal seedling. Once new growth begins conditions the cup/sleeve is usually filled with following grafting, plants are roots within 4-6 weeks (Fig. 16b). When placed in darkness for 7-14 days until the new shoot reaches about 400 mm in length. t at Fig. 15 Scraper used to prepare the etiolated section of the stem for treatment with a 1% IBA/0.5% NAA formulation. Fig. 16 In the rootstock cloning method of Hofshi (1996) the (a) application of a 'clear plastic 240 ml cup' filled with a rooting medium placed on a pretreated portion of the etiolated shoot to induce rooting; and (b) root development in the cup after 4-6 weeks and the subsequent grafting to a scion variety and detachment from the 'nurse' seed.

15 12 sufficient top growth has been made the scion cultivar is grafted and when mm of new growth has been made the rooted, clonal tree is removed from the nurse seed. The grafted, rooted cutting still in the clear container, is placed on the heated propagation table in the greenhouse for an additional 4-6 weeks to harden-off. Subsequently it is replanted into a large nursery bag and transferred outside and grown on to the required size for field planting. This takes an additional 8-12 months. Following detachment of the rooted, grafted cutting from the "nurse seed" a second cloned rootstock can be produced. However, in this case the cloned rootstock may not be grafted to the scion variety until after removal from the "nurse seed". This latter stage depends on the residual vigour of the "nurse seed" and the prevailing climatic conditions (time of year). Following propagation of the second clonal rootstock, the "nurse seed" can be used as a standard seedling rootstock. Prior to our visit ACW had mistakenly etiolated seedlings and then promoted rooting using the same procedure as that for cloning elite rootstock. On inspection of this material it was noted that the IBA/NAA-treated, etiolated section of the seedlings had produced strong, uniform roots which were superior to the rooted sections of elite rootstocks that were being produced using the same technique (Fig. 17). This comparison clearly illustrated the differences in response between physiologically juvenile and mature plant tissues with respect to their rooting ability. Research into growth manipulation using different spectral light composition during nursery production was beginning at the time of the visit. Investigations in Israel have shown that with ornamental species different spectral light ^L Fig. 17 Comparison in the rooting of etiolated stems of physiologically mature (left) and juvenile (right) plant tissue (ACW Nursery, California). composition, modified by using coloured shade cloths, can have a marked effect on growth. Red and Pearl shade cloth (manufactured by Polysack in Israel) have been found to promote vegetative growth while Grey enhances side branching in some ornamental species. Results of this research will be made available by ACW when the experiment has been completed. Greenleaf Nursery (California) Greenleaf Nursery is located at Bonsall in southern California and is owned by Dave Brodersen and Wendy Katze (wendv@gleafhurserv.corn). The nursery is relatively new to avocado production and is targeting avocado trees on cloned rootstocks in the 2002/03 production season. The owners believe that production of cloned rootstocks where the "nurse seed" has been removed will give them the most reliable product to sell to their clients and to achieve this they are using a similar production technique to that seen at Allesbeste and ACW nurseries. Greenleaf Nursery has put its own interpretation on cloned rootstock production with some modifications of the Allesbeste/ACW production methods. There was little difference in the production of "nurse seedlings" between Greenleaf and ACW in that they were both using polystyrene drinking cups as their nursery pot.

16 13 Greenleaf has moved away from using 'Zutano' as a "nurse seed" and were using 'Lula' imported from Florida. They claimed greater success in producing cloned rootstocks when using 'Lula'. The sourcing of nurse seed for production of nursery trees is a major flaw in the Californian production of nursery trees due to the unknown status re: Avocado sunblotch viroid. a *. Fig. 18 Inserting clear containers over the etiolated stems prior to filling with Greenleaf was using clear drinking containers (16 fluid ounces) with a central hole punched in the base as a container to hold the potting media around the etiolated shoot (Fig. 18). The etiolated stem was pre-treated prior to potting by scraping either side of the shoot for 4-5 cm and then sprayed with 1% IB A/0.5% NAA. The most notable innovation of Greenleaf Nursery is the changed pot dimensions and shape for finishing off plants prior to sale. They are using pots with a vertical ribbing utilising forestry research which has demonstrated that this type of pot design produces trees with vertically straight root systems reducing the potential of the roots to curl and the plant to become root-bound (Fig 19). Fig. 19 Vertically-ribbed, 5 L pots used by Greenleaf Nursery to grow plants through to saleable size (left). The pots were held vertical by a frame and base that grouped them into packs of six which facilitated handling in the nursery. An inverted plant shows the straight root system produced by the vertically-ribbed pots (right). Huerto California Nursery (Chile) Huerto California is one of several nurseries producing avocado trees for the industry in Chile and is owned by the Maghdal family (cmagdahl@safex.cl). The Chilean avocado industry has grown rapidly over the past 10 years with currently ha planted and expansion still occurring at the rate of about 1000 ha year" 1. The industry is entirely based on seedling rootstocks with about 90% of these being 'Mexicola', which is highly susceptible to Phytophthora root rot. While disease pressure in Chile is relatively low the tree population is relatively young with many orchards being planted on heavy clayey soils which could create problems in the future (Fig. 20).

17 14 mm^t.4i:>^: '*' '.' /.»KKvv>. ~ P: ffh&ji 'Nabal' is also used as a seedling rootstock in some situations where water quality is poor and salinity levels are high as this rootstock has greater tolerance to salt than 'Mexicola'. A small orchard of 'Hass' grafted to seedling 'Velvick' trees has also been planted and the trees are performing very well despite the cool, Mediterranean climate. Fig. 20 ' *». Although disease pressure is low in Chile Phytophthora root rot is beginning to appear in 6-7 year-old orchards. superphosphate. The soil is steam sterilised using the standard pasteurisation methodology. Seeds are extracted from the fruit and scarified clipping the tops of the cotyledons prior planting to improve the uniformity germination. The floor of the nursery concreted and copper fungicide footbaths are placed at strategic entry points to ensure traffic into the nursery does not carry disease. Trees are grown on wooden slats about 100 mm off the floor level (Fig. 21). This would not comply with ANVAS standards. The nursery is covered by black 35% shade cloth but research is intended to investigate different light spectrum modified The Huerto California nursery uses a potting mixture of 70% composted pine bark; 15% coarse sand; and 15% soil (described as clayey). One kg of S m" 3 of potting mix is added to lower the ph along with 180 g of Fig. 21 Grafted 'Hass' trees on 'Mexicola' seedling rootstocks at Huerto California. Note the wooden slats supporting the plants. by coloured shade cloths. The nursery trees are irrigated using 1.2 L h" 1 micro-tube with a min daily watering period. Nutrition is also delivered via the irrigation system with plants receiving 1 g of urea, 0.5 g of KN0 3 and 0.2 cc of phosphoric acid plant" 1 month" 1. Huerto California uses a grafting technique different to that commonly in practice by other nurseries visited. Scions are sidegrafted into trees about 150 mm above the soil level (Fig. 22). About 10 days after grafting the tree is headed back to four leaves above the graft which encourages the grafted scion to grow. The claim is that trees grafted this way grow quicker as they have a greater photosynthate base with Fig. 22 Side grafting into the trunk of young nursery trees (left). The tree is headed back to four leaves above the scion about 10 days after grafting when the buds begin to swell (right). the leaves retained above the graft.

18 15 The trunk extending above the graft is also used to support the growth from the new scion. Huerto California has begun a cloned rootstock programme initially with 'Duke 7'. The nursery is currently using the Brokaw technique and has about 2000 trees on cloned 'Duke T ready to plant in spring 'Zutano' is currently used as the "nurse seed" and etiolation is carried out in an air-conditioned dark room with the temperature held at around 25 C (Fig. 23). Plants in the etiolation room are sprayed with a mixture of copper oxychloride (300 g L" 1 ) and Benlate 50 (70 g L" ) to control fungal pathogens and are normally in the room for days. Fig. 23 Plants in etiolation California an air-conditioned, room at Huerto Fig. 24 Constriction ring placed on the plant to encourage rooting and to sever the clonefromthe "nurse seed". On removal from the etiolation room a metal washer is placed over and moved to the base of the etiolated growth (Fig. 24) and immediately above the washer the bark is slit vertically and painted with 0.75% IBA. The treated area is covered with potting media to encourage the development of roots. Experience to date suggests that it will take months to produce a saleable tree on a cloned rootstock. Grafted trees on seedling rootstocks sell for $6 AUD and it is expected that grafted trees on cloned rootstocks will sell for about $10 AUD. California Rootstock Status of Countries Visited The Californian industry developed largely on seedling 'Topa Topa' (Mexican) and to a lesser extent 'Mexicola' (Mexican) and 'Lula' (Guatemalan/West Indian) rootstocks. 'Topa Topa' and 'Mexicola' left a legacy of rootstocks highly susceptible to Phytophthora root rot while 'Lula' gave significant rootstock/scion incompatibility as seen in the plantings of the Rancho California development near Temecula. In more recent times the industry has produced trees on cloned rootstocks of which 'Duke 7' has been the main one used. New lines such as 'Thomas' and 'Merensky 2' are being introduced to the industry as clones however, there is still a significant percentage of trees still propagated on seedling 'Topa Topa'. Rootstock breeding programme Dr John Menge, University of California Riverside, has a rootstock breeding programme aimed at producing rootstocks with greater tolerance to Phytophthora root rot than material currently available to industry. To facilitate this programme rootstock lines with proven Phytophthora tolerance have been planted in isolation to reduce the risk of introduction of foreign pollen.

19 16 Crosses being made are UC2001 (seedling 'Duke 7'), 'D9' (an irradiated seedling of 'Duke 7'), 'Thomas', and 'Spencer' ('Spencer' looks like 'Sharwil' but has high anise level in the leaves - grows with a crawling habit exhibiting loss of apical dominance. Growth habit may be due to greater sensitivity to boron deficiency). In a second breeding plot 'Spencer', 'Toro Canyon', 'PP4' ('Zentmyer') and 'Thomas' are planted. To date the greatest number of seedlings with Phytophthora root rot tolerance has come from 'Thomas' where it is used as the maternal source. Cross pollination is encouraged by netting trees and introducing bees to the enclosure (Fig. 25). Where netting is not used the breeding plots are at distance from other avocado trees so as to prevent the introduction of other pollen. For root rot evaluation seeds are collected from the various maternal sources, germinated in shallow boxes in a vermiculite rooting media and inoculated with Phytophthora cinnamomi. Roots ; & are examined for decay and then the best survivors potted on into larger pots where they are rechallenged with Phytophthora cinnamomi. The survivors showing promise are then taken into the Fig. 25 Caged trees in the California Phytophthora tolerant rootstock breeding programme. field ^ rown on ] P duce propagation material for cloning trials where the candidate rootstock is grafted to 'Hass'. The Menge rootstock breeding program is using "Brokaw" technique developed for 'Duke 7' with a clear plastic sleeve so that rooting can be monitored. 'Topa Topa' is used as the "nurse seed". It has been found that some seedlings are quite difficult to clone using the 'Duke 7' protocol. This result is consistent with similar comments made by other nurseries visited that were cloning other rootstocks of interest. Florida The Florida avocado industry is based on seedling rootstocks with 'Waldin' the most commonly used. 'Lula' has also been used to a lesser extent. There is no data available to support the choice of 'Waldin' as a rootstock and it seems that its use is largely through convenience (fresh seed available at the beginning of the nursery production cycle), availability of large numbers of seeds and easy to work with in the nursery. New Zealand The New Zealand industry is based on seedling rootstocks with 'Zutano' the predominate variety used. According to the genealogical relationships among cultivated avocado as determined by RFLP analysis, 'Zutano' is a Mexican/Guatemalan hybrid but most closely related to the Mexican race (Davis et al., 1998). Examination of orchards revealed a significant percentage of rootstocks showing an appreciable overgrowth of the scion (Fig. 26). It was also observed that many of the fruiting 'Hass' trees had cropped on limbs that had not refoliated during the summer. It is suspected that these trees would have badly defoliated during

20 17 flowering which is usually what happens when the degree of overgrowth shown in Fig. 26 is present. I* IV> ' v*-* fcl w /4\. Fig. 26 'Hass' grafted to 'Zutano' seedling rootstock at Sweetwater, Kaitaia. Note the strong overgrowth of the scion. No reason could be given for the use of seedling 'Zutano' by the New Zealand industry but it is most likely due to that of availability and ease of production of nursery trees. 'Hass' has also been grafted to seedling 'Reed' and planted in some orchards. Grower comment was that they believed the 'Hass'/'Reed' combination was more productive than the 'Hass'/'Zutano' combination but no reliable records were available. New Zealand soils are highly leached and deficient in boron hence this nutrient must be carefully managed. Boron uptake and translocation to the top of the tree is difficult where 'Zutano' is used as a rootstock, thus reflecting its Mexican inheritance. On the other hand where 'Reed' has been used as a rootstock grafted to 'Hass' boron uptake is significantly more efficient and growers unaware of the differences have had phytotoxicity problems where they have treated both rootstocks at the rate required to raise levels in trees with seedling 'Zutano' rootstocks. This result clearly shows the necessity of knowing the parentage of rootstocks used and managing their nutrition in an appropriate manner. The fact that 'Zutano' is responding similarly to 'Duke 6' and 'Duke 7' rootstocks with respect to boron uptake might also imply that other nutrients are similarly effected and therefore this combination may also be having a negative impact on postharvest fruit quality. Elite performing 'Hass' trees were inspected at Ron Bailey's property at Te Puke where it was claimed that trees were grafted to seedling 'Fuerte'. However, the trees showed no sign of incompatibility at the graft union which almost always occurs when this combination is used in Australia. South Africa Through to the mid-1980's the South Africa avocado industry was widely planted on seedling 'Edranol' rootstocks. With the development of 'Duke 7' as a Phytophthora-tolerant rootstock the industry swung strongly across to cloning this rootstock along with 'Martin Grande' (the G755 series). In good soils 'Marin Grande' was soon abandoned due to high vigour and poor cropping however, it still has a role in heavy, wet soils where vigour is limited and yields improve (Note: a.**\ * * 5M-w.t, Fig. 27 Phytophthora trunk canker in a 'Duke T rootstock in South Africa.

21 18 similar experience with this 'Marin Grande has been reported from Western Australia and the rootstock should be re-evaluated under these conditions). It is worth noting that in the Phytophthora rootstock evaluation block at Westfalia Estates, 'Edranol' cloned as a rootstock and grafted to 'Hass' is a highly susceptible rootstock with most trees dying in the first 18 months after planting. 'Duke 7' is still the mostly widely cloned rootstock used in South Africa but is susceptible to Phytophthora trunk canker (Fig. 27) and due to its Mexican origin has difficulty in translocating boron to the scion variety. Rootstock breeding programme Merensky Technological Services has a rootstock breeding programme aimed at producing lines with greater Phytophthora root rot tolerance and is run on similar lines to the Californian (Menge) programme. A mother seed block has been established in isolation and includes 'Duke 7', 'Thomas' and 'Merensky 1' and 'Merensky 2'. These latter two rootstocks were recovered as escape trees from orchards severely affected by root rot. The mother seed block has trees of the breeding lines planted in close proximity that produce seeds via open pollination. About 3000 seeds are produced each year and these are germinated in vermiculite and challenged with Phytophthora cinnamomi. It is normal to get only two new candidates each year to take forward for further field testing. Note: both the Californian and South African Phytophthora root rot tolerant breeding programmes are based on Mexican race lines. Research in subtropical Australia to date indicates that when Mexican race rootstocks are grafted to 'Hass' there is deterioration in postharvest fruit quality (e.g. 'Duke 6', 'Duke 7' and 'PI'). As the results have been sourced from a limited population it is too early to know if all Mexican race rootstocks will give this result but sufficient evidence is available (Willingham et al. 2001; Marques et al. 2002) to suggest that new Mexican race rootstocks be tested thoroughly for all production aspects prior to release to industry. The Merensky Technological Services rootstock improvement programme is also recovering rootstocks from high-yielding trees for cloning and re-evaluating performance when grafted to 'Hass' (Fig. 28). Fig. 28 A consistently high-producing tree identified at Westfalia Estates. Rootstock recovery is proceeding via inducing rootstock sprouting from wounds made below the graft union.

22 19 References Davis, J., Henderson, D., Kobayashi, M., Clegg, M.T. and Clegg, M.T. (1998) Genealogical relationships among cultivated avocado as revealed through RFLP analyses. The Journal of Heredity 89(4), Marques, J.R., Hofman, P.J. and Wearing, A. (2002) Rootstock influences 'Hass' avocado fruit quality and fruit minerals. Journal of Horticultural Science and Biotechnology (in press). Willingham, S.L., Pegg, K.G., Cooke, A.W., Coates, L.M., Langdon, P.W.B. and Dean, J.R. (2001) Rootstock influences postharvest anthracnose development in 'Hass' avocado. Australian Journal of Agricultural Research 52, Apogee Plant Growth Regulators There is currently interest in developing a new growth regulator, Apogee for use on avocados. Apogee has been registered to suppress vegetative growth in apples and other temperate fruit crops in Europe and the USA. It is claimed that the product has a reduced environmental profile when compared to the triazoles currently use by the avocado industry, e.g. paclobutrazol (Cultar ) and uniconazole (Sunny ). Like the triazoles, Apogee effects the biosynthesis of gibberellic acid in plants but the pathway through which it acts is thought to be different to the other products. Research with Apogee on avocados is currently being carried out in California, South Africa and Chile. With apple registration it is more likely that an Apogee registration is likely to happen for avocados in California providing efficacy can be shown while those countries exporting to markets (Europe and the USA) which do not have triazole registration see this product as a more acceptable chemical for use on avocados. Apogee is a Japanese product (calcium 3-oxido-5-0x0-4-propionylcyclohex-3-enecarboxylate) and has been commercialised by the BASF Corporation, PO Box 13528, Research Triangle Park, NC27709, USA. Preliminary results from Chile have shown that when Apogee was foliar applied to avocados at full bloom at 125 g 100 L" 1 water with a non-ionic surfactant (Break ) the product increased yield (Table 2). However, when applied at 250 g 100 L" 1 of water the yield was significantly reduced possibly due to phytotoxicity. A smaller crop load resulted in larger fruit. Table 2 Effect of mid-bloom sprays of Apogee on fruit size and yield of 'Hass' avocado growing in Chile. Data are mean values of 4 trees. The planting density of the orchard was 278 trees ha" 1 and the results reported are for the 2000/2001 fruiting season (Results courtesy of Sociedad Gardiazabal y Magdahl Ltda). Treatments* Yield (kg tree" 1 ) Fruit number tree" 1 Mean fruit size (g) Projected yield (t ha" 1 ) Control Apogee at 125 g 100 L" Apogee at 250 g 100 L" Apogee was applied with the non-ionic surfactant Break. To date the use of Apogee to retard summer shoot growth has not been successful even though

23 20 application rates as high as 400 g 100 L" 1 of water have been used. Application of Apogee is dose sensitive with herbicide effects occurring when too high a volume of the 125g 100 L" 1 formulation was applied to trees in mid-bloom in California (Fig. 29). Fig. 29 An excessive application of Apogee to 'Hass' at mid-bloom in California resulted in reduced leaf size and leaf burn (a) with stimulation of axillary buds in new shoot growth (b). Yield data is not available but from visual inspection there is no apparent increase in fruit set in this experiment. Naphthalene-acetic acid (NAA) Auxins have been shown to control shoot regrowth on avocado stumps following top-working trees. Regrowth below the graft union when top-working avocado stumps to new cultivars will compete with new scion growth. Hand removal of water-shoots is labour-intensive and expensive. In California it has been shown that shoot regrowth can be controlled by spraying stumps with either a 1% ethyl ester or sodium salt formulation of naphthalene acetic acid (NAA) in a 30% aqueous solution of a white acrylic paint when shoots were mm long. Shoots wilted 2 h after treatment and were killed by both NAA formulations 45 days after treatment without any deleterious effect on the newly grafted scions that were shielded at the time of treatment. 'SmSAAJ^a-.' ' - >, K ltlftf- (ft*** * ;,^% Fig. 30 The central leader of 'Reed' trees headed back and treated with 1% NAA formulated in 1:1 acrylic paint/water. Shoot regrowth is controlled to about 15 cm below the treated area (California, 2002). i,4 In California similar NAA technology was being applied experimentally to 'Reed' trees trained to central leaders and planted at ultrahigh density. A 1% NAA formulation made with 1:1 acrylic paint and water was being applied to control tree height after the central leader had been headed back (Fig. 30). The treatment was effectively controlling shoot regrowth for up to 18 months following treatment. This technology warrants investigation under Australian conditions where major cuts are made to trees during canopy management operations.

24 21 Canopy Management South Africa Most of the avocado production in South Africa is in areas that have a summer-wet subtropical climate that encourages vigorous vegetative growth. Research on canopy management has been carried out for a number of years by Dr P. Stassen, Institute for Tropical and Subtropical Crops, Nelspruit in the direction of intensively training 'Hass' trees to central leader form (Fig. 31). During our time in South Africa we saw no evidence that the industry had accepted the results from Dr Stassen's research programme. Instead, canopy management was being practiced along the lines developed in Israel where trees were being strategically pruned with mechanical saws and the regrowth treated Fig. 31 Training 'Hass' trees to central leader with the growth retardant, Sunny. form at the Burgeshall Research Station, Hazyview. Canopy management as practiced by Westfalia Estates involves pruning trees twice each year: in the winter as soon as the crop has been harvest the trees are given a light prune (usually by hand) to remove any material that falls outside the main shape of the tree and again in early summer with mechanical saws to establish the fruiting platform for next season. The trees are treated with % Sunny at mid-bloom to encourage increased fruit retention and larger fruit. The summer regrowth is also Sunny -treated when about 100 mm of shoot growth has occurred using 0.3% Sunny plus a follow-up with 0.2% Sunny 2-3 weeks later if required. For major re-shaping of trees the alternate sides are mechanically pruned in consecutive years so as to retain at least one productive fruiting face on the trees. In this instance the heavy cut is made in the winter as soon as the crop has been harvested. Canopy management was also discussed with Andre Ernst of Allesbeste Nursery who also has a large avocado orchard. Andre is able to maintain t/ha on pruned 'Hass' orchards that are planted at 5 x 7 m while still keeping trees to a manageable size. It was also claimed that the pruning strategy followed reduces the intensity of alternate bearing. The following steps in pruning for this orchard were described as follows: 1. the severe prune when starting the process is given to only one side of the hedgerow in the first year with the canopy being cut back to within 2 m of the trunk immediately following harvest (about July). The second side of the hedgerow is similarly pruned the following year immediately following harvest (Fig. 32). 2. The pruned side of the hedgerow is summerpruned (December/January) and the regrowth treated with Sunny when regrowth is about mm long. It is usual to give two Sunny sprays about two weeks apart using 0.25% each time. The second spray may be Fig. 32 Pruned and Sunny -treated Hass Trees at Andre Ernst, Tzaneen, SA.

25 22 omitted if regrowth is sufficiently controlled with the first spray and cold weather occurs early enough to stop growth. Agral is added to the Sunny at 0.01%. It has been noticed that zinc leaf concentrations have fallen to lower levels in the pruned blocks compared to the unpruned blocks of trees. Hall and Sons, Tzaneen had crowded, overgrown orchards which they were just beginning to rejuvenate. Interestingly, they were being advised by Dr Nigel Wolstenholme who had recommended that they follow the Anderson selective limb removal canopy management strategy so that crowding could be reduced whilst still retaining orchard productivity (Fig. 33). Although in the early stages of bringing the orchards back under control the manager of the programme was satisfied Fig. 33 Selective limb removal being practiced for with the results being achieved. canopy control at Hall & Sons, Tzaneen. Florida The Florida avocado industry suffered severe damage following Hurricane Andrew which struck in the summer of The lesson learnt from this storm was that trees that topped to about 5 m or less suffered less damage than trees that were allowed to grow taller. While topping in most orchards is still practiced there is no consistency between farms on the overall management of orchards. Selective limb removal was being practiced by one grower who was feeling his way with the technique and already claiming that his orchard was now more productive than when he was hedging both sides of the trees. There were many orchards where the side canopies were poorly lit and fruiting had retreated to the tops of the trees. In these cases growers seemed unwilling to make a decision to prune. The overall impression was there was little to be learnt with respect to canopy management in Florida. California The Californian industry is the home of close planting followed by progressive tree removal until finally, when the remaining trees become too large, the orchard is stumped and allowed to regrow. While in theory this option can work well, in practice it does not always work out as growers become reluctant to make timely removal of trees. There was some evidence that this system is still being practiced by some growers but there are also many orchards where crowding has occurred and the fruiting has moved to the tops of the trees. Ultra-high Density Orchards Ultra-high density avocado orchards are currently being promoted in California by the Chairman of the CAC R&D committee, Mr Reuben Hofshi. Several orchards have been planted in different localities in the state and have clocked up impressive production figures. The concept involves planting varieties with natural central leader architecture such as 'Reed' and 'Lamb Hass' at 2.25 x 2.25 m spacing (1973 trees ha" 1 ). Trees are annually pruned, usually after the current season's

26 23 crop has set, to ensure that they do not occupy more than their allocated space in the orchard. Pruning is aimed at maintaining tree shape, height, light interception and orchard access while ensuring productivity is maintained. Tree height is maintained at a level that allows hand harvest of all fruit from the ground and is assisted by the use of NAA applications to the central leader which is topped at about 2.3 m from the ground (see NAA in Plant Growth Regulators) (Fig. 34). Using the protocols Fig. 34 An ultra-high density orchard of 'Reed' in its 5 th year developed by Hofshi a six-yearold 'Reed' orchard has yielded where it produced 65 t ha". 6.5,26.0, 65, and ha" in the third to sixth years from planting, respectively. Current limitations to this production system are market acceptability of cultivars with suitable architectural structure for close planting, e.g. 'Gwen', 'Lamb Hass', and 'Reed', and the cost/return of this production system based on the life of the orchard which has not yet been practically quantified. The is no doubt that this work is pushing the production system to its limit and the with some modification the practices used by Hofshi may have value for the Australian industry for the production of 'Reed' and 'Lamb Hass'. Avocado Sunblotch Viroid (ASBV) Avocado sunblotch disease was first recorded in California and has subsequently been reported in most avocado growing areas including Australia, Florida, Israel, Peru, South Africa, Spain and Venezuela. It is possible that the disease is more widespread because of symptom variability and the presence of symptomless carriers. Although initially thought to be caused by a virus due to its graft transmissibility, the disease was later shown to be caused by a viroid. V-" The symptoms of avocado sunblotch disease vary considerably and are influenced by cultivar, environmental conditions and viroid strain. Typical symptoms usually include stem streaks, lesions and discoloration of the fruit, and a variety of foliar symptoms (Fig. 35). Importantly, some infected trees remain symptomless. The most consistent symptom of sunblotch infection is the appearance of yellow, orange or white streaks on the stem and petioles. These streaked areas often darken Fig. 35 Avocado sunblotch symptoms on the twigs, leaves and fruit of 'Hass'.

27 24 and become depressed. Foliar symptoms usually appear as either chlorotic zones associated with vascular tissues, which commonly appear bleached, or as a general variegation of white, yellow or pink areas. In some cases, both symptom types are present on the one leaf. Leaves may also be deformed. Fruit produced from infected trees usually develop sunken white, yellow or red blotches or streaks and are usually small, deformed and unmarketable (Fig. 35). The bark can have a rectangular cracked appearance, often referred to as 'crocodile skin' or 'alligator hide'. The symptoms of sunblotch may be irregularly distributed throughout an infected tree. Severely affected trees are stunted and develop a low, sprawling habit that leads to increased exposure and injury from the sun. Fruit set and yield is also usually reduced. Variants of ASBV of between 246 and 251 nucleotides have also been detected in diseased trees and four types have been identified in Australia. A discussion at the Bundaberg conference with Dr John Menge of the University of California prior to the study trip revealed that when variants are present in trees there may be up to a 10% reduction in yield. He also indicated that infection by variants is likely through natural root-grafting between trees. When visiting Florida we spent time with Dr David Kuhn who was based at the USDA-ARS National Germplasm repository near Miami. Sixteen ASBV variants have been identified in avocado trees by this group and it was emphasised that they are unstable viroid particles which readily mutated. Advice was received that there should be zero tolerance of ASBV or its variants in commercial avocado orchards. The tissue culture unit at the University of Florida, Homestead is currently working on eradicating ASBV and its variants from infected material. The basic technique is to culture meristems from infected material on the assumption that the cells have divided and grown faster than the viroid can invade them. These 'shoot' meristems with elongate but to date root initiation from this type of material has not been successful. To rescue this material whole embryos excised from seeds and carrying both root and shoot primordia are placed in tissue culture and allowed to develop into new rooted plants. When of sufficient size the shoots grown from meristems are micro-grafted onto the small trees and eventually the new plants established in a greenhouse. If successful this technique may be used to recover ASBV-free material of valuable varieties/rootstocks. New Zealand has been recently re-examining the viroid status of its industry using the Queensland University of Technology analytical services. It has been discovered that the same four variants found in Australian material are also present in New Zealand trees. Advice from New Zealand virology experts is that the NZ avocado industry should adopt a zero-tolerance policy on ASBV or its variants in commercial orchards. During the course of their investigation some doubt has been thrown on the reliability of analytical services provided by QUT in that split samples have tested both positive and negative {Note: in a more recent conversation with Dr Rob Harding of QUT it was disclosed that the source of cross-contamination had been identified and steps taken to ensure that it would not be a further problem. Dr Harding gave assurances on the integrity of the ASBV analyses being carried out for the Australian industry). New Zealand has recently sent coded material to Dr Barry Manicom of the ARC- Institute for Tropical and Subtropical Crops at Nelspruit in South Africa for testing. An indication of the reliability of this laboratory should be known by September, Our knowledge on ASBV has steadily increased as new, more sensitive detection methods have been developed. With current knowledge it is recommended at this point in time that a zero-

28 25 tolerance policy should be adopted for all trees being propagated and sold from ANVAS nurseries as soon as practicable and this be enforced through the routine sampling of leaves for ASBV analysis from nursery trees. This policy should be supported by rigorous testing of nuclear and multiplication blocks and that the Australian industry should work closely with the New Zealand industry in identifying a reliable and affordable indexing service. Recent literature pertaining to ASBV is attached in Appendix X. Packaging and Marketing Avocados Returnable Plastic Containers (RPC's) are the future for most horticultural products; they are the present system for some firms marketing avocados from South Africa to England (Fig. 36). They offer many benefits over traditional methods for all sections of the marketing system. To the farmer there is a cost saving, for transport there is the strength and reliability, to the importer/distributor improved ventilation for cooling and ripening and for the retailer no disposal problem. Some of these benefits apply to all sections of the marketing chain and increasingly consumers are becoming aware of environmental issues which are minimised through the elimination of fibre-board and polystyrene containers. Fig. 36 Returnable plastic crates being Fig. 37 Well-ventilated RPC's used by used for pre-wrapped products in Westfalia Estates, South Africa for Tescos, London. ripening avocados. Packaging in South Africa Westfalia estates in South Africa ripen and distribute avocados to supermarkets nationally using RPC's. When the fruit has been treated and graded it is placed into well-ventilated containers, which allow full exposure of the fruit to ethylene gas while ensuring even temperature control. These particularly well-ventilated containers are used only in the ripening process (Fig. 37). The sprung or ripened fruit is pre-packaged into smaller containers which are used in many European supermarkets, or packed directly into RPC's. The benefit of the latter for transporting produce is that they can be fitted into each other for return to the pack-house, whereas those used for the ripening process do not "compact stack". The process for distributing avocados in RPC's to the national supermarkets is very similar to England and is explained below. However South Africa does have some other rather unique methods of containerizing avocados for transport. One of these is to package the avocado (even Fuerte) into 20 kg onion sacks which may be stacked up to 2 m high on open trucks and transported over long distances. These are generally sold in wet markets and roadside stalls. A roadside stall can consist of those we see in Australia down to one person sitting on the side of the road with a single sack of avocados. Another container in South

29 26 Africa is made from fluted plastic with solid plastic corners. This container is being used for citrus and bananas and will soon be used in the avocado industry. There is a world wide-patent on this container. Preparing and Exporting Sportac is not accepted on all international markets and with the demand for organically grown produce increasing in Europe Westfalia has developed a postharvest treatment for anthracnose and stem-end rot, which passes the requirements for organic registration. The product is Biocoat and it is made up with a blend of glycols and alcohols and is sold as a fruit coating and not a fungicide. When used in the field, immediately fruit is cut from the tree the stem-end is dipped in Biocoat. A further application is applied in the pack-house where the fruit is treated with a solution containing Biocoat and Avoshine (Fig. 38). The demand for organic produce is strong from Europe. Most Supermarkets are demanding a proportion of their avocados come from registered organic groves. Although the prices for that product are the same as traditionally grown produce the incentive is otherwise traditionally grown avocados will not be purchased. Fig. 38 Coating avocados with Avoshine (wax) (left) and Biocoat (fungicide), which are applied on the grading and packing line (right). Westfalia produces 2.5 million cartons of avocados of which 11.4% are organically grown. This is becoming increasingly successful; at present the No.l Pack out from traditional avocados is 80% and organic 70%. English supermarkets have increasingly developed a workplace health and safety ethical component to their quality assurance program. They are increasingly under scrutiny by lobby groups of various political persuasions to ensure that they are not sourcing food and products from third world countries where labour/children are being exploited. Hence, South Africa avocado growers routinely are inspected by delegations from the UK to ensure they are complying with the mandatory requirements. The RPC Trail in the UK Avocados are packed into IFCO RPC's for export: note again the good ventilation on sides and bottom. IFCO crates are handled similar to the CHEP pallet system in Australia, where the user pays a rental fee, and in this instance the crates are picked up from a depot in South Africa and when used by the importer in England returned to the South African depot. These containers fold

30 27 flat to minimise transport cost when not packed, they fold out to hold 4 kg of fruit (Fig. 39). With avocados the importer who features here is Mack Multiples, a large family-owned firm who import produce from all over the world to resell to all the major supermarkets. Mack Multiples have separated the different types of produce into divisions under control of specialist supervisors. Avocados come under the division of salads, whose technical manager is Dr. E.R. Moorhouse. Individual products in turn have their specialist procurer: for avocados it is Ms Hana Weston. Fig. 39 The collapsible IFCO RPC being used in South Africa. The RPC folds flat (left) for storage and return for reuse and is simply and quickly assembled when required (right). The major countries supplying avocados to Mack Multiples are South Africa, Mexico, Israel and Spain. With respect to fruit quality Israel and Spain present no problems. This can be for two reasons: their proximity to the European market; and the other their Mediterranean climate being less conducive to anthracnose and stem-end rot. However, in some years both South Africa and Mexico can have serious post-harvest problems reducing consumer confidence in the product. Currently Mack Multiples is looking at other countries which can supply during May to October, for example Peru is being investigated. Australia would be considered provided the price was right, and our quality could be guaranteed. If Australia was to export it should only be from orchards using rootstocks and fertilizer programs with proven records for fruit quality. to the cool room Fig. 41 Ripening room at Mack facilities at Mack Multiples, UK. Fig. 42 Firmness tester at Multiples, UK. Mack Multiples, UK. Overseas containers arriving at Mack Multiples are sealed onto the cool room facilities (Fig. 40) and from here they are transferred to the cool rooms and kept at 5 C until being placed in the

31 28 ripening room (Fig. 41). Here they are subjected to temperatures between C. Ethylene is rarely used as they consider it has a bad effect on shelf-life. However, ethylene facilities are available if needed to speed up the ripening process. The ripening fruit is monitored with a firmness tester (Fig. 42) each day from four different levels from each pallet. After firmness testing each of the four fruit is dissected for inspection. The problems that show at this stage are disease, maturity and cold damage. As mentioned earlier, anthracnose and stem-end rot can occur from particular countries some years, while pulp graying due to extended cold storage can be quite a major problem. Braising does not tend to show up at this stage but expresses Latter in fruit which are kept back from each consignment. They believe the bruising problems are occurring during their handling of the sprung fruit, especially when placing it into prepacks. Quoting from Mack's "England is the most difficult country to market avocados as their customers insist on ripened fruit whereas the rest of Europe generally handles hard fruit". If Australia was to market in the UK there are aspects of our cooling and ripening chain we could help with and they would be receptive to our ideas. Mack Multiples use an in-line non-destructive firmness tester (Fig. 43) to grade different stages of ripeness. However, they experience similar problems with accuracy as expressed in other countries. Fig. 43 (a) An in-line firmness tester used by Mack Multiples in the UK; (b) the fruit is tapped a number of times as it is rotated past a probe; (c) the packing line delivers fruit of uniform firmness. The method this machine uses is to tap fruit a number of times as they are rotated past sensors (Fig. 43b). However different horticultural practices and growing conditions produce fruit of different qualities, such as skin thickness and elasticity (e.g. the more copper fungicide applied during growth the more brittle the skin). Also the seed size alters the way the fruit will respond to this method. Constant calibration of the machine is necessary. Computer software on the packing line is designed to use the grader to deliver fruit of a uniform firmness, not size as usually is the case (Fig 43c). Fruit showing no signs of being sprung is returned to the ripening room. When pallets of avocados ripen evenly they can then be packed into smaller containers (Fig. 44) without the added cost of using the firmness-testing machine, here are two different sizes the two pack and the four pack (Fig. 44). Many different sized packs are used and are still generally distributed in RPC's

32 29 =M"'r - ««. * ft Fig. 44 Packing fruit into small tray packs. Fig. 45 Two and four fruit per pack are the most commonly used counts. This three pack (Fig. 46) is especially good for riper fruit as they are well protected from damage, even in the shopping trolley. The sticker is added at the wholesalers with variety, weight, price, country of origin and use-by-date (Fig. 47). These pre-packs are generally used for Fig. 46 The three pack used for riper fruit. Fig. 47 Fruit in tray packs are comprehensively labelled. the smaller fruit however, the general preference by consumers is the larger singles (Fig. 48). The reasons for this being they are perceived as better value and no matter what produce managers do people still want to feel the fruit which is easier to do when not packaged. Fig. 48 Most fruit going to the supermarkets are sent out in these RPC's however, to the outlets with smaller turnovers small fibre-board containers are used. The containers are palletised and loaded onto refrigerated transport. From the day they arrive in the ships containers to the day they are distributed they are under controlled temperature conditions. Fig. 49 shows a pallet of RPC's ready for distribution. For hygiene purposes all staff and visitors have to wear caps and gowns, while all Large singles that are not j e w e l l e r y must be remov ed and hands thoroughly prepacked are still preterred by..., customers.

33 30 Fig. 49 A pallet of avocados packing in RPC's ready for distribution to a retail outlet. Chain Store Displays in California The displays in the fresh produce section of the major Californian chain stores were arguably the best seen on the study tour. In each case where we visited the section managers were knowledgeable and enthusiastic about they produce they were selling. The fruit was of high quality and spoiled product quickly removed from the displays. In most cases displays were either colour coordinated (Fig. 50), arranged by product e.g. all the various kinds of potatoes (European and sweet), or complementary products. The latter was most usually done for avocado which was invariably displayed with tomatoes (Fig. 51). The contrast between the bright red and Fig. SO Bananas displayed in the fresh produce section of a large chain store in Los Angeles. green of the fruit was eye-catching and according to the section managers drew customers and assisted in moving fruit. The avocado/tomato stand also included packaged ingredients with instructions for making fresh salsa's and guacamole along wim corn chips thus promoting the theme of fresh, healthy savoury dips/snacks for various occasions. Many of the ideas in the display came from the Californian Avocado Commission who also supported the stands with dietary information on avocados that accentuated the health benefits of the fruit. Fig. 51 Avocados commonly displayed with tomatoes in Californian chain stores. The stand also provides other ingredients for making salsa's and euacamole.

34 31 Marketing in New Zealand The New Zealand industry has an export focus due to the volume of production and the limited size of the domestic market. Fruit are currently exported to Australia, California and Japan. Fruit is shipped overseas through exporters who source fruit from a group of growers committed to servicing that particular exporter. Team Horticulture is a good example of an exporter and sources fruit from about 300 growers. The company employs an agronomist/consultant who provides technical advice on production issues and also coordinates harvesting and shipping during the fruit picking season. With the collapse of the Californian avocado market following the September 11 destruction of the Twin Towers in New York the New Zealand industry faced a dilemma of having to sell more fruit on to the Australian market. This posed a risk of collapsing the Australian price. Cooperation between all NZ exporters in the volumes they were sending on a weekly basis to Australia resulted in the sale of 30% more fruit than they had ever sent before at record returns back to growers. The strict flow-plan imposed went right back to grower level with set weekly volumes given to individual suppliers. This managed market supply prevented excess fruit accumulating in the supply chain and maintained wholesale prices. There are potential lessons to be learnt by the Australian industry from this operation. However, there are greater difficulties on the domestic scene as there are many individuals marketing fruit and collusion between suppliers may not be favourably viewed by the ACCC.

35 32 Contacts/People Visited on the Study Tour Arpaia, Dr Mary Lu, University of California, Kearney Agricultural Center 9240 S. Riverbend Ave. PARLIER, CALIFORNIA 93648, USA (Phone: ; arpaia(a),uckac.edu). Bender, Dr Gary. University of California Cooperative Extension, 5555 Overland Avenue, Building 4, SAN DIEGO, CALIFORNIA , USA( Blanden, Bill. Koeltehof Farms, Box 69, KIEPERSOL 1241, SOUTH AFRICA Bower, Dr John. Horticultural Science, School of Agricultural Sciences & Agribusiness, Private Bag X 01, Scottsville 3209, South Africa (Tel: ; Fax: ; boweri (gtnu.ac.za). Brodersen, Dave. Greenleaf Nursery, via Puerta del Sol, BONSALL, CALIFORNIA 92003, USA (Phone: ; Fax: ; wendv@,gleafnurserv.com). Chao, Dr Thomas. Department of Botany and Plant Sciences, 2137 Batchelor Hall, RIVERSIDE, CALIFORNIA , USA (Phone: ; Fax: ; ctchao(a>citrus.ucr.edu). Crane, Dr Jonathan. Tropical Research and Education Center, Institute of Food and Agricultural Sciences, SW 280 Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; ihcr@mail.ifas.ufl.edu). Cutting, Dr Jonathan, Avocado Industry Commission, PO Box Bethlehem, TAURANGA, NEW ZEALAND (Phone: ; Fax: ; ionathancutting(q),xtra.co.nz). Donkin, Derek. Technical Manager, SAAGA, PO Box 866, TZANEEN 0850 SOUTH AFRICA (Phone: +27 (015) /7; Fax: +27 (015) ; Cell phone: ; saaga@pixie.co.za; Web: Erasmus, Gerhard. Allesbeste Nursery, PO Box 91, TZANEEN 0850, SOUTH AFRICA (Phone; ; Fax ; ahernst(o),mweb.co.za). Ernst, Dr Andre. Allesbeste Nursery, PO Box 91, TZANEEN 0850, SOUTH AFRICA (Phone; ; Fax ; ahernst(o).mweb.co.za). Faber, Erica. Nursery Manager, Merensky Technological Services, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; ericaf@hansmerensky.co.za). Hearne, Rodney. Danroc Sub-tropical Fruit and Nuts, PO Box 24, KIEPERSOL 1241, SOUTH AFRICA (Phone ; Fax ; danroc@iafrica.com)

36 33 Hofshi, Reuben, Del Rey Avocado Company, 1260 S. Main St., FALLBROOK, CALIFORNIA 92028, USA (Phone: ; Katz, Wendy. Greenleaf Nursery, via Puerta del Sol, BONSALL, CALIFORNIA 92003, USA (Phone: ; Fax: ; Kohne, Dr Stefan. Merensky Technological Services, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; co. za). Knight, Hedley. Born Free Farm, PO Box 295, KIEPERSOL 1241 SOUTH AFRICA (Phone: +27 (013) ; Fax: +27 (013) ; Cell phone: ; Kremer- Kohne, Sylvie. Merensky Technological Services, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; Kuhn, Dr David. Department of Biological Sciences, Florida International University, University Park, MIAMI, FLORIDA 33199, USA (Phone: ; Fax ; Litz, Dr Richard, Tropical Research and Education Center, Institute of Food and Agricultural Sciences, SW 280 Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; McKee, Brandon. Department of Plant Pathology, University of California, RIVERSIDE, CALIFORNIA , USA (Phone: ; Fax: ; McMillan, Dr Bob Tropical Research and Education Center Institute of Food and Agricultural Sciences SW 280 St. Homestead FL , USA. (Phone +1 (305) ext. 270; Fax: +1 (305) ); Moon, Pamela. Tropical Research and Education Center, Institute of Food and Agricultural Sciences, SW 280Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; Moorhouse, E.R. Technical Manager, Salads MACK MULIPLES DIVISION Tranfesa Road, Paddock Wood. KENT TN12 6UT LONDON, UK (Phone: +44 (01892) ; Fax: +44 (01892) ; Cell phone: ; Philcox, Mark. Grove Services Inc. of Miami, PO Box , HOMESTEAD, FLORIDA , USA (Phone: ; Fax: ; net). Partridge, Colin. Agronomist/consultant, Team Horticulture, PO Box 13445, TAURANGA, NEW ZEALAND (Phone: ; Fax: ; colin@teamhort.co.nz). Ploetz, Dr Randy. Tropical Research and Education Center, Institute of Food and Agricultural

37 34 Sciences, SW 280Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; Porritt, Roy. Box 69, KIEPERSOL 1241, SOUTH AFRICA Retief, Wouter. SAAGA, PO Box 866, TZANEEN 0850, SOUTH AFRICA (Phone: Cell phone: ; Web: Roe, Dennis. Westfalia Estates, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; Schaffer, Dr Bruce. Tropical Research and Education Center, Institute of Food and Agricultural Sciences, SW 280 Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; Smith, Duncan, Westfalia Estates, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; Snijder, Bram. Burgershall Research Station, Institute for Tropical and Subtropical Crops, Agricultural Research Council, Private Bag X501, KIEPERSOL 1241, SOUTH AFRICA (Phone: ; Fax: ; Suarez, Dr Isidro, Tropical Research and Education Center, Institute of Food and Agricultural Sciences, SW 280Street, HOMESTEAD, FLORIDA 33031, USA (Phone ; Vorster, Louis. Westfalia Estates, PO Box 14, DUIWELSKLOOF 0835, SOUTH AFRICA (Phone: ; Fax: ; Weston, Hana. Avocado Procurement Executive, MACK MULIPLES DIVISION Tranfesa Road, Paddock Wood, KENT TN12 6UT LONDON, UK. (Phone: (01892) ; Whyte, Alan. Spring Field Farms, PO Box 1547, LOUIS TRICHARDT 0920, SOUTH AFRICA (Phone: +27 (015) ; Fax: +27 (015) ; Cell phone: ; Witney, Guy. Production Program Research Manager, California Avocado Commission, 1251 E. Dyer Road, Suite 210, SANTA ANA, CALIFORNIA 92705, USA (Phone: ; Fax ; Wolstenholme, Dr Nigel. 30 Wavell Drive, Wembley, PIETERMARITZBURG 3200, SOUTH AFRICA (Phone: ;

38 Appendix I Sequence Diversity among Avocado Sunblotch Viroids Isolated from Single Avocado Trees

39 Sequence Diversity among Avocado Sunblotch Viroids Isolated from Single Avocado Trees R.J. Schnell, l D.R Kuhn, 2 C.T. Olano 1 and W.E. Quintanilla 1 in previous studies we developed an avocado sunblotch viroid (ASBVd) specific reverse iranscriptkm-poiymerase chainreaction (RT-PCR) protocol and tested all avocado germplasm accessions in the U.S. National Germplasm Repository (NGR) for ASBVd. It was found that 19% of the accessions were infected with the viroid. To characterize the population of sequence variants, single trees of seven accessions with a history of infection were selected. RT-PCR products were cloned and sequenced from these trees and a totai of 60 sequence variants were identified from i22 clones. Most of the variants were found lo be unique to the individual tree from which they were isolated. The fraction of unique variants was 81% in 'lima Late', 81% in 'Aycock Red', 67% in 'Hawaii', and 60% in 'Young Special". Analysis of the sequence data suggests that variants evolved within s treeratherthan arising from multiple infection events, supporting the quasispecies concept for ASBVd. Changes in therightterminal loop (RTL), previously associated with leaf variegation and bleaching, were found in all variants isolated from bleached tissue; however, 16 clones isolated in nonsymptomatic tissue also had changes in the RTZ.. The addition of a U between bases 115 and 118 was found in 2! of the 23 clones isolated from the bleached tissue but not found in the 99 clones isolated from green tissue. These data suggest that the insertion of a U between bases 115 and 118 is responsible for the bleaching condition. KEY WORDS: Viroids; capillary electrophoresis; sequence variants; quasispecies. INTRODUCTION Sunblotch, an economically important disease in avocado, is associated with a small, covalently closed, circular, single-stranded RNA known as avocado sunblotch viroid (ASBVd) (6). Viroids age, RNA replicons that cause transmissible diseases in many economically important crop species; they have been separated into two families based on structure and sequence similarity. The largest family is the Pospiviroiaae, which is further divided into five genera. This family contains the potato spindle tuber viroid (PSTVd) types. The second family, the Avsunviroidae, has two monotypic generti, ihe Avsuniviroid containing ASBVd, and the Pelamoviwid. containing peach latent mosaic viroid (PLMVd) (14). The Avsunviroidae differ from the Pospiviroidae in that die former can form a hammerhead structure and have the ability to self-cleave in vitro sad in vivo (4). ASBVd was the first characterized member of ihe Avsunviroidae. It is a persistent, often undetected, pathogen of Persea amerkana Mill., the common avocado. Received May ;receivedinfinalform Aug.!3, 2001; posting Oct. io l USDA-ARS. National Germplasm Repository. Subtropical Horticulture Research Station. Miami, Ft USA lc-mail: miare@ars-grin.gov]. 2 Dept. of Biological Sciences, Florida international University. Miami. FL 3319?, USA. Phytoparasitica 29:5,

40 A highly sensitive and accurate detection protocol mat utilizes reverse transcription and the polymerase chain reaction (RT-PCR) has been developed to selectively amplify ASBVd from partially purified RNA extracts (10). ASBVd-specific RT-PCR has been used to screen all the accessions in the USDA germplasm collection for avocado at the National Germplasm Repository (NGR) in Miami, Florida. Of the 429 trees in the collection, 81 were found to be infected (8). Initial banding patterns indicated that sequence variants were present among the viroid populations. It is generally recognized that naturally occurring.viroid infections often contain more than one sequence variant in a single host plant. Rakowski and Symons (7) identified 16 sequence variants isolated from three plants and Semancik and Szychowski (11) found three variant types that were associated with differential symptom development. In this study, we investigated the sequence diversity occurring in seven different avocado trees by sequencing 122 clones yielding 60 variants. We were interested in determining whether ASBVd populations consist of predominantly a few wild type variants, as has been reported for PSTVd (S), or, alternatively, if they exist as a quasispecies as found for other RNA pathogens (3). Additionally, we isolated ASBVd from symptomatic and non-symptomatic tissue to determine if differences previously reported between variants isolated from the two types of tissue in California (11) were also occurring among the accessions in the germplasm collection. MATERIALS AND METHODS Plant materials Tissue samples were collectedfromfive plants maintained at the NGR in Miami, Florida ('Lima Late*, 'Aycock Red', 'Gainesville', 'Wilson Popenoe', and 'Young Special') and two plants ('Arguello' and 'Hawaii') maintained at the Tropical Agricultural Research Station in Mayaguez, Puerto Rico (Table 1). Avocado trees selected for this study had visually expressed disease symptoms on fruit at Miami and tested positive for ASBVd viroid when assayed with ASBVd-specific RT-PCR (8). Samples were collected from a single tree for each accession. Leaves and flowers from different sides of the trees were sampled. All tissue sampled was non-symptomatic except for bleached leaf tissue of Aycock Red. RNA extraction, cdna synthesis, cloning, and sequencing RNA was extracted from closed flowers or older leaves using Amresco Rapid RNA Purification Kit (Amresco, Inc.; Solon, OH, USA). cdna was synthesized using the primers and methods developed by Schnell et al. (10). Amplified cdna products were fractionated by gel electrophoresis in 13% molecular grade agarose (Sigma; St Louis, MO, USA) in 0.5 x TBE and visualized with ethidium bromide. The band at approximately 2S0 bp was excised. DNA was isolated from the gel slice using QIAquick Gel Extraction Kit (QIAGEN Inc.; Chatsworth, CA, USA) and quantified with a Dynaquant 200 (Amersham Pharmacia Biotech Inc.; Piscataway, NJ, USA.) Purified DNA was cloned using a TA Cloning Kit (Invitrogeir, San Diego, CA, USA>. Plasmid isolation was accomplished using Wizard Plus SV Minipreps DNA Purification System (Promega; Madison, WI, USA) following the kit protocol. Plasmid DNA was quantified fluorometrically. Sequencing was accomplished by use of the ABI Prism Dye Terminator Cycle Sequencing Ready Reaction Kit with Amplitaq DNA Polymerase FS (Applied Biosy stems; Foster City, CA, USA). This kit employs rhodamine chemistry. Seven hundred ng of 452 R. J. Schnell eial.

41 TABLE 1. Avocado accessions used for ASBVd isolation and characterization Miami Accession Race 1 Source Plant Clones Number Unique no. date sequenced or variants variants Lima G Lancetilla Late Honduras Aycock W.I. Lancetilla, Red Honduras Gainesville M Gainesville Florida 3757 Wilson W.I. Tela Popenoe Honduras Young M Mulberry Special Florida Arguello W.i. Managua, Nicaragua 3893 Hawaii W.l. Oahu. Hawaii G. Guatemalan; M. Mexican; W.I.. West Indian. template from each clone was amplified separately with universal primers Ml3 forward and M13 reverse-(promega). Unincorporated dye terminators were removed by use of Centriflex gel nitration cartridges (EDGE Biotechnologies; Gaithersberg, MD, USA). Dye Terminator Cycle Sequencing products were analyzed on an ABI Prism 310 Genetic Analyzer (Applied Biosystems). Data were collected using the rapid sequencing run module. Electrophoresis was performed using POP6 polymer and capillaries 47 cm in length and 50 /im internal diameter. Analysis was performed using the ABI Prism DNA sequencing analysis software (Version 3.0) with CE-2 basecaller. For each clone both strands were sequenced and assembled. Sequence data analysis Sequences were aligned manually in Sequencher 3.0 (Gene Codes Corp. Inc.; Ann Arbor, MI, USA). The text output from Sequencher 3.0 was converted into a Nexus file and edited in MacClade 3.07 (Sinauer Associates; Sunderland, MA, USA). The aligned data yield a composite sequence of the variants, including all insertions, of 252 bases. Insertion/deletion events (indels) were obtained by aligning sequences in the most parsimonious way. Six dummy characters were included for the indels. Of the 258 aligned characters, 212 characters were constant and 27 variable characters were parsimony-uninformative. Seven of these 19 parsimony informative characters were due to indels. One such indel was an 8 nucleotide deletion in three of the variants (CF18, CF40 and CF44). Since such a deletion was likely a one-time event and no intermediate forms of the deletion were detected, this indel was represented by a dummy character, preventing overvaluation of the deletion. All other deletions or insertions were treated similarly. Thus, the final set of parsimony-informative characters contained 19 characters, where 12 were substitutions and 7 were indels. RESULTS Sunblotch symptoms on the fruit have been detected on selected trees of Lima Late, Aycock Red, Gainesville, Wilson Popenoe, and Young Special for a number of years; however, only Aycock Red has produced leaf tissue with bleached areas. Symptoms have not been detected on Arguello or Hawaii. Phyloparasitica 29:

42 ISO I,1 1»l J 1 1 M, MA *.*-! i 1 1 AycodcRw! fe 1 i» v a; on G c 6 * I» i «m 1 i GAA [c a»%-mkx»c6aia G* oxn*cooaem GAC cue oco /v t ce taaaeasmx COG GDC Ma-igGcy GMUSG coc cot p~6 CCC t «90 i el i A C II U G o n a c u A» / U U GftG CCA GD-GKCKft GGRG CS 'ISO I 60 A c «a d 1 1* 1 c «o u c u u & aj u < GO-GRGMA. GGMS CG in ncntd cboc occ G AA, oi Fig. 1. Nucleotide variation,* frequency of changes and number of varianu with these changes in the trees of 'Lima Late' (panel'a), 'Aycock Red' (panel b), Hawaii' (panel c) and 'Arguello', Wilson Popenoe' and 'Young Special* (panel d), located.on the proposed secondary structure of variant J ('Parsimony informative changes are shaded.) In this study 122 clones were sequenced, yielding 60 sequence variants. Two of the variants (J02020 and M31085) had been reported previously (7,13). Sixteen variants were obtained more than once from different RT reactions. These included J02020 and M31085 and 14 variants that were evaluated and reported in a Single Strand Conformational 454 R.J. Schnell el al.

43 Polymorphism (SSCP) variant discrimination study (9) (GenBank accession f AF through AF229828). It is likely that the remaining 44, which were cloned only once, are true sequence variants, and not arttfactual products of the RT reaction, as variation did not occur randomly throughout the genome but rather at specific sites (Table 2). The variants were given die designation of Chapman Field (CF) 1 through 58 beginning with the 14 variants found in more than one RT reaction. The 44 new variants (CF15-CF58) have been deposited in GenBank, accession numbers AF through AF Changes giving rise to die 19 parsimony-informative characters among die 60 variants occurred in the Right Terminal Loop (Rrz.). in the Left Terminal Loop (LTL) and in the stem area. In the LTL four informative changes were detected at bases 3,-5,7 and In the stem area 8 informative changes were detected at bases 45,58,62,67,146,164, 183 and 205. Three variants, CF18, CF40 and CF44, had significant deletions that could involve either bases adjacent to the cleavage site or bases 66-73, as they are tandem repeats. In the Rrt informative changes were detected atfivesites, 113, , , 124 and Changes at the parsimony informative sites are illustrated using tinted boxes in Figure 1 along with other changes found in the sequence variants. Twenty-one variants were sequenced from 38 clones isolated from Aycock Red. Three of these variants were also found in Lima Late and one in Lima Late and Hawaii. The remaining 17 variants were unique to mis tree. The variants were isolated from bleached and green sections of the leaves. Fifteen variants (23 clones) were isolated from die bleached tissue and six (15 clones) were isolated from the green tissue. Using the classification system of Semancik and Szychowski (11), variants can be classified as Bleached (B), Variegated (V) or Symptomless Carrier (SC) types based on the sequence. The most frequent changes detected from the bleached tissue were an addition of a U between 115 and 118 and me addition of A or AA between 122 and 128 in the RTL. We did not detect UU to AA exchanges at or a transition from A to U at 119, as reported by Semancik and Szychowski (11). Of die 15 variants isolatedfromaycock Red bleached tissues-all but two had die additional U or UU at From Lima Late, 31 sequence variants were characterizedfrom56 clones. One of these was die wild type variant J02020 and one was die previously described sequence M Of the remaining 29 sequences, three were also found in Aycock Red and one in Aycock Red and Hawaii. The remaining 25 variants were unique to lima Late. Nine variants were isolated from 11 clones from die accession Hawaii. One of die variants was identical to a variant found in Lima Late and Aycock Red, anotiier was identical to a variant found in Arguello and a third was identical to a variant found in Wilson Popenoe. The six other variants were unique to Hawaii; three had changes in die RTL- Of the five variants isolated from 12 clones in Young Special, one was identical to and another was identical to M The three ouier variants were unique to Young Special. Two of die diree variants had changes in die RTL- Two variants were found in Arguello, with one identical to a variant found in Hawaii and die ouier unique to Arguello. Neither variant had changes in the RTL- A single variant was found in Wilson Popenoe that was identical to a variant found in Hawaii and a single variant found in Gainesville was identical to J02020 (Fig. 1). Parsimony analysis of die variants was inconclusive, due to die large number of variants and small number of parsimony informative characters. The parsimony informative changes were summarized bytissuetype, bleached or symptomless (Table 3). In addition, the number of times a particular mutation was found in a clone was reported. The frequencies of the mutations were reported because die viroids appear to be more like populations than like lineages. Phytoparasitica 29:5,

44 TABLE 2. Sequence variant (SV). avocado accession, isolate tissue (IT) and nucleotide changes (NC) obtained from seven avocado trees for variants sequenced from a single RT reaction (one identical cdna clone in each case except C.T7, which had thee. Superscript ~ indicales bleached leaves.) -=- v? Residue RE TS^ ft Residue HE

45 TABLE 3. Sequence heterogeneity at parsimony informative sites among ASBVd variants associated with different symptoms Residue no. in SB-l z Nucleotide Symptomless No. of Bleached No. of changes 2 Carrier clones of clones of (SC) SC with bleached change with ^ change 'Wild type' - None + 13/99 Left Terminal Loop 3 U-*A + 51/99 + 7/23 5 U-*A + 17/99 + 7/23 7 -A + 21/ U + 7/99 Right Terminal Loop 113 A-»G + 2/ Uor+UU + 21/ U + 6/ A-»G + 1/99 + 1/ A + 9/ / AA + 4/99 + 3/23 Other 45 G-»A + 2/99 62 G-»A + 3/99 58 U-»C + 5/ UUCCGACU + 1/99 + 2/23 67 U-+G + 14/99 + 1/ A-»G + 1/99 + 9/ U->C + 2/ G-+A + 2/ A-»G + 1/99 + 1/23 *The residue number and-nucleotide exchange areas described for ASBVd SB-1 (13). DISCUSSION The first report of sunblotch occurring in Florida is from the mid 1930s and the source of infection was believed to be budwood obtained from California (12). Some of this budwood was used for propagation at die Plant Introduction Station at Chapman Field, now the NGR at Miami. Another introduction of ASBVd to the NGR is thought to have occurred when infected budwood was brought from Lancetilla, Honduras, to Miami in 1968 (R. Knight, personal communication). All of the trees used in this study tested positive for ASBVd in the 19% survey and five of the seven have had symptoms for at least the last 3 years (8). Infecting indicator seedlings in a temperature-controlled greenhouse produces disease symptoms in one year and within 2 years symptoms are expressed under ambient conditions (1). Thus, the viroid has had ample time to complete many replication cycles in these plants. In this study isolates yielding significant numbers of cdna clones (Lima Late, Aycock Red, Hawaii, and Young Special) were obtained from field-grown trees. The viroid isolates were found to be complex mixtures of sequence variants. Similar findings were reported by Rakowski and Symons (7). Six sequence variants were identified from ten cdna clones in one isolate and in another isolate ten variants were identified from 17 cdna clones. In Phytoparasitica 29:5,

46 contrast, an isolate obtained from greenhouse-grown graft-inoculated seedlings contained two sequence variants in the 23 cdna clones sequenced, the predominant sequence being that of the inoculum, J02020 (83%). The variants detected in this study could be classified using the system proposed by Semancik and Szychowski (11), whereby changes associated with the bleaching condition were correlated with changes in the RTL- Our results are in agreement with these observations. Of the 23 clones sequenced from Aycock Red bleached tissue, 21 had the addition of a U or UU between bases 1 IS and 118 and 18 had additions of an A or AA between bases 122 and 128. We also detected changes in the Rxr. in one variant isolated from green tissues of Aycock Red, nine variants isolated from Lima Late, three variants isolated from Young Special, and two variants isolated from Hawaii. This suggests that among all variants contained in non-symptomatic tissue of an individual tree, a significant number will have these changes in the Rrz, - Of the 99 clones isolated from green tissue, 15 had changes in the RTL; however, not a single variant isolated from the 99 clones in green tissue had the insertion of the U or UU between bases 115 and 118 (Table 3). Most of the trees studied in this investigation have had sunblotch symptoms on the fruit for a number of years, but only Aycock Red has had the bleaching leaf symptoms. Variant M31085 was reported as a representative sequence of the variegated population (11), yet it was found in symptomless tissue in this study. The results presented here, indicate that individual trees mat have been infected for many years will contain multiple ASBVd sequence variants. Some of these variants will contain changes in the RTL «n bom non-symptomatic and bleached tissue. The change in sequence that is associated with the bleaching condition is the addition of a U or UU between bases 1 IS and 118, not changes at bases as was reported previously. In fact, of the three variants found in the bleached tissue by Semancik and Szychowski (11), two had the inserted. U between 115 and 118. Semancik and Szychowski (11) hypothesized that certain variants would cause bleaching in tissue even if they could not be detected and that an enlargement of the right terminal loop was an important aspect of the viroid structure that caused bleaching. We did not find the clone reported previously to cause bleaching among 38 clones sequenced from Aycock Red. We performed SSCP analysis on CF1-CF14, J02020 and M31085 and found the changes in the RTL such as the addition of U in bases and A in had no effect whatsoever on single-stranded conformation. The biggest changes in conformation were due to substitutions in die Lri- In fact the Aycock Red variants from the bleached tissue could not be distinguished from J02020 by SSCP (9). We are currently preparing clones of viroid dimers containing the U insertion in the region to determine if this single nucleotide change is sufficient to cause bleaching symptoms in the leaves of any avocado cultivar. Using mutated PSTVd and infecting plants widi several variants simultaneously, Owens et al. (S) observed that while some mutants were stable in the plant, many reverted to the wild type. It has been reported that, in individual plants multiple sequence variants are present early in the infection cycle. Over time, the wild type variants reproduce preferentially until they predominate in the population (2). Our results are antithetical to these observations. Lima Late contained 31 of the 60 variants reported here, 25 unique to this individual. Aycock Red contained 21 variants, 17 unique to Aycock Red, while Hawaii contained nine variants, six unique to Hawaii. 458 R.J. Schnell et al.

47 The quasispecies concept essentially defines a host/parasite relationship, where in an individual host large populations - often called swarms - of sequence variants of a given virus or viroid pathogen occur. Variants within the quasispecies population compete during replication and are subjected to natural selection. If environmental conditions remain constant, these wild type variants will continue to predominate; however, if environmental conditions change, then selection may favor one of the variant forms over the wild types (3). Lima Late and Aycock Red have been infected with ASBVd for many years, providing ample opportunity for the establishment of a few predominant types. However, the number of sequence variants found in these two trees does not support the predominance of a few wild type variants. We believe that our data support the characterization of these ASBVd populations as quasispecies. The two previously reported sequences, J02020 and M31085, were found in Lima Late and Young Special; J02020 was the only variant found in Gainesville. These two variants were not found in Aycock Red, Wilson Popenoe, Arguello or Hawaii. Not detecting the prototypic J02020 or M31085 in these four plants could be due to (a) the original infecting viroid being lost, or (b) founder effects. In any case, 81% of the variants isolated from Aycock Red and Lima Late, and 67% of the variants isolated from Hawaii were unique to these trees. Indeed, certain mutations at particular positions were found to be quite common in all trees, but the assemblage of such mutations in a single variant showed a great deal of diversity. If variants were simply due to a high rate of mutation and the survival of neutral or advantageous mutations by selection, then the diversity of variants should be high but essentially similar from tree to tree. Instead, we see that each tree has its own unique set of variants, made up of unique assemblages of the most common sites for mutations. This would be the likely outcome of a high mutation rate and stochastic sampling, which would be expected for quasispecies swarms. The frequency of any variant in the population does not appear to be linked to the frequency of the immediate ancestors nor does it appear to predict the appearance of variants due to accumulation of further mutations. Rather, the frequency of any variant or its appearance in future generations may depend on which variants in the population are replicated. The opportunity to be replicated may depend on environmental and other chance factors, such as the genotype of the host. In this study the bleaching variants occurred exclusively in Aycock Red. The Aycock Red tree is located 7 m from the Lima Late tree, and therefore environmental factors (soil and weather) may be considered constant. Using parsimony analysis of the variants' sequence, we could not detect any evolutionary pattern that depended on the accumulation of mutations. The evolutionary trajectory of these variants appears to be based on stochastic processes and population disequilibria, making it impossible to conclude anything about the evolution of a single variant or its effect OR future populations of the viroid. REFERENCES 1. DeGraca. J.V. and Van Vuuren. S.P. (1981) Use of high temperature to increase therate or avocado sunblotch symptom development in indicator seedlings. Plant Dis. 65: Diener.T.O. (19%) Origin and Evolution ofviroids and Viroid-Kke Satellite RNAs. VirusGenes 11: Duarte. E.A.. Novella. I.S.. Weaver, S.C.. Domingo, E., Wain-Hobson, S.. Clarke. D.K. a al. (1994) RNA virus quasispecies: significance for viral disease and epidemiology. Infect Agents Dis. 3: Navarro. B. and Flores, R. (1997) Chrysanthemum chlorotic mottle vtroid: Unusual structural properties of a subgroup of self-cleaving viroids with hammerhead ribozymes. Pmc. Natl. Acad. Sci. U.S.A. 94: Phytoparasitica 29:

48 5. Owens. R.A., Hammond. R.W.. Gardner, R.C., Kiefer, M.C.. Thompson, S.M. and Cress. D.B. (1986) Sitespecific mutagenesis of potato spindle tuber viroid cdna: Alterations within premelting region 2 that abolish infectivity. Plant Mol. Biol. 6: Palukaitis. P., Hatta, T.. Alexander, D.McE. and Symons. R.H. (1979) Characterization of a viroid associated with avocado sunblotch disease. Virology 99: Rakowski. A.G. and Symons. R.H. (1989) Comparative sequences studies of variants of avocado sunblotch viroid. Vimlogy 173: Ronntng. CM.. Schnell. R J. and Kuhn, O.N. (1996) Detection of Avocado Sunblotch Viroid and estimation of infection among accessions in the National Germplasm Collection for avocado. Pmc. Fla. State Horiic. Soc. 109: Schnell, RJ.. Kuhn. D.N. and Olano. CT. (2001) Detection of avocado sunblotch viroid variants using fluorescent single-strand conformational polymorphism analysis. Electrophoresis 22: Schnell, RJ., Kuhn, D.N.. Ronning, CM. and Harkins. D. (1997) Application of RT-PCR for indexing avocado sunblotch viroid. Plant Da. 81: Semancik. J.S. and Szychowski, J.A. (1994) Avocado sunblotch disease: a persistent viroid infection in which variants are associated with differential symptoms./ Gen. WroJ. 75: Stevens. HE. and Piper, R.B. (1941) Avocado Diseases in Florida. U.S. Dep. Agric. Circ Symons. R.H., (1981) Avocado sunblotch viroid: primary sequence and proposed secondary structure. Nucleic Acids Res. 9: Van Regenmonel, M.H.. Fauquet, CM., Bishop, D.H.L.. Carsten. E.B- Estes. M.K.. Lemon. S.M. et al. (2000) Virus Taxonomy: The Seventh Report of the International Commitiee on Taxonomy of Viruses. Academic Press Inc., New York, NY. 460 RJ. Schnell era/.

49 m Appendix II Genealogical Relationships Among Cultivated Avocado as Revealed Through RFLP Analyses

50 Genealogical Relationships Among Cultivated Avocado as Revealed Through RFLP Analyses J. Davis, D. Henderson, M. Kobayashi, Matthew T. Ctegg, and Michael T. Ctegg AnottymoiisDfvAfragroerrtsfToml^ nana MHL) were cloned into a ptasmid vector and used to screen a total of 36 curfvars. There is a high level of potymorpmsrn among cukhrars allowing ad cmthewsto be assigned a unique genotype based on 14 genetic lod. A cluster analysis of genetic stntilariries among culnvars revealed three major dusters that corresporidtotl*e three major ladal grovpm^ appear to reflect culnvars derived through irrterrac'ial hybridization. The cultivated avocado (Persea amenama) is comprised of three botanical rac with certainty because the mature avoca fruit where the pollen parent is known es: P. americana van americana (West Indian race), P. americana van drymifblia flowers but fewer than 1,009 are destined do tree produces in excess of 1,000,000 (Schect & Cham) Make (Mexican race), to yield maturefruitmost fruit drop prior and P. americana 'van gaertemalensis Was to maturity, leaving a smalt residual of suc- (Guatemalan race) (Bergh et al 1973; Popenoe 1941). The origins aad domestication of the various races is obscure, sod Dsh both the parentage of individual trees, Genetic markers provide a way to estabin fact die appellation west Indian is a misnomer because avocado was introduced fer the likely genealogical histories of par and through retrospective analyses, to in to the West indies from Central America ticular culuvars (Mhameed et al 1997). In by the early Spanish settlers (PopenoeW, an initial attempt to explore avocado relationships using generic markers,, we in Zentmeyer G, and Scheiber G. unpublished manuscript). Many of the named cuttrvars vestigated a number of culdvars and members of other closely related Persea spe of avocado are selections from open pollinated progenies and the pollen parent is cies (P. noccosa Mez, P. nubigena L. Wats, often unknown or a matter of speculation. P. steyermarknc.k. Allen, and/! shiedeana In some cases, the open pollinated selections that yielded successful cutttvars are roplast DNA probes (Furnier et al. 1990). Nees) using rdna, nuclear gene, and chlo- believed to have arisen from interracial These data revealed that all three races crosses. Thus, for example, the adnvar are distinguished by unique mutations Hass, the predominant commercial cultivar in CaBfornia was first obtained as an pothesis that the Guatemalan race origi and the data provided support for the hy open pollinated seedling of unknown parentage (Seetye 1994). Current speculation nabigena and P. steyermarkil Furnier etal nated through hybridization between P. among avocado breeders (Bergh R and (1990) concluded that P. americana could From the Department ol Botany and Plant Sciences, Martin G, personal communicatjon) is that be expanded to include P.ttoccosa,P. nubigena, and P. steyermarkii as one large University of Catuorma. Riverside. Cataorala 92SZ1. cv. Hass is derived from a Guatemalan x This work was supported in part by the CauTorota Mexican hybrid that was then backcrossed to a Guatemalan pollen parent. It recommendations of Bergh et al 0973) Avocado Commission. Vfc thank Dr. Richard Whtdms species winch concurs with the earlier for comments on the manuscript and fortnspbotograr^irslnlk.um Francis lor hhastlslanre In Sndlng is now possible to test this and similar based on the analysis of terpene patterns. a photogenic nee. Dr. Richard Utz and Cray Martin speculations regarding genealogical relationships among avocado cntthiars using for hejpbig obtain avocado materials, and Dr. Michael m this article we expand our analyses of Cumoimp (or the drawing of the genealogy. We woi*t also Uke to thank Gray Maitm aad Bob Bergh for information on putative relationships among cuutsars ed avocado through the use of anonymous genealogical relationships within cultivat molecular markers. and for their assistance locating the avocado materials used m this study. Address correspondence lo The traditional bases for inferring parentage have been morphological or were relationships is of more than academic in RFLP markers. Knowledge of genealogical Dr. Michael T. Ctegg at the address above or cleg5vmcracl.ucr.edu. based on the proximity of a potential pollen source. It Is very difficult to produce lationships is fundamental to the design terest; accurate information on genetic re 1996 The American Genetic Association 8* of 319

51 Table 1. The genmy M! for each cnltl»ar based on Ure e»m{dm<^m WW ainu 9JKI WW i own 1 IMO V4KI 80RI 95RI non U7R 121RI 364R 366R) 26RV 50RV 53RV 64ARV 64BRV 73R\ Ame Zt Bkmdo CoUinrcd Pollock Cellons Anaheim Esther Nabal H Reed Daily Unda U NimSoh Fuerte u Noga HX Qiaen Hass H K2S Bacon Zntano Lyon TWlIe HnkertoD WhltseU RiOCOD league Duks Thomas ii tgnacio u Gamer Yama Meaioola Duke IbpaTopa n Khan R schtendeam G75SC u G755A ii Each column contains the data for a specific RFLP probe. u n efficient plant improvement strategies and such knowledge is essential for effective germplasm conservation. The data from this study reveal a high degree of polymorphism for anonymous nuclear DNA fragments, confirming the impression Oat cultivated avocado possesses a rich and diverse gene pool. Ouster analysis among a set of 36 cultivars, based on a measure of genetic similarity, reveal groupings that conform with racial designations and that appear to distinguish cuttrvars derived from interracial hybrids. Thus RHP analyses appear to provide a powerful basis for interring genetic relationships among cultivated avocados. Materials and Methods Plant Materials Collections of avocado germplasm ate maintained at the University of California South Coast Field Station and on the campus of the University of California, Riverside. Because the method of avocado propagation is clonal, based on bud grafting. each cultivar represents a single genotype. Table 1 provides a list of the cultivars surveyed and their source. In addition, >?sdnerfeana was inctuded as an outgroup in the cluster analyses. DNA Preparation DNA was firrrarted from 10 g of newly emerged leaf tissiie using a modification of die protocol of Rawson et al 0982)- The leaf tissue was homogenized in 50 ml of grinding buffer Q.QB mlf Tris, 25 mm EDTA, 035 M sucrose, 50 mm KCL 5» polyvinylpyrrolidone. 10 mm diethyldithiocarbainic add, 03% mercaptoethanoi) using a small Waring Mender. The homogenate wasfiltered through cheesecloth and centrifuged at 12,000 G for 20 min (4 C). The pellet was resuspended m 6 ml of rysiiig buffer QOOmM EDTA. 50 mmtrifrhcl ph % Triton X-100, 256 sarkosyl. 50 ug/ml Proteinase K) and incubated at 37"C in a shaking incubator for 2 h. The lysate was centrifuged at 15J0OO G for 10 min (t*c) and the supernatant was precipitated with 2/3 volume isopropanol at -20 C for 30 min. The precipitate was pelleted at G for 15 min. (/TQ. The peflet was resuspended in TE buffer (10 mm Tris-HCl, 1 mm EDTA ph 8.0) and the DNA purified through a single CsQ gradient (if = ) as described by Rawson et al (1982). The DNA sample was precipitated. washed with 70% ethanol, dried under vacuum, and resuspended in TE buffer at a concentration of 1 ug/ul Ckming Anonymous DNA Fragments Total DNA from cv. Hass was digested with the restriction enzyme Ps& and size fractionated on 1% agarose gels, (ftfl was used to select against chloroplast DNA because it cuts the cmoroplast genome infrequently into relatively large fragments.) The fraction corresponding to approximately 400 bp in size was eluted following standard procedures. The size fractionated DNA was Ggated into a puc18 vector and transformed into coff strain JM101 following established procedures (Sambrook et al 1989). Recombinant colonies were selected and minipreps of DNA were screened to verify insert size, and these were further analyzed via Southern transfers to exclude any chloroplast-derived in- 320 The Journal ol Heredity 1998:89(4)

52 Table 1. Extended 74W SOiW 95RV tlfjrv untv KIBJf 364RV 38ERV 260 SOB S3III 64AIB 131 7«««lion 1I7HI I2in 364DI 3a It SS II It SS ~ II SS II u II SS II 22 S II II It SS II U a II It II II SS It It S It II II II It tl U It u II 11 II» II II It S It II B II SS u It 11 u U 11 U It II U II It ! II it It a II n II n sens. Clones with appropriate inserts were screened against Southern transfers of DNAfromeach of the three races of avocado to establish whether the resulting RFLPs were both single-copy and polymorphic. Finally, selected clones were used as probes to an Fj progeny derived from self-fertilized Hass to verify nudear inheritance. Soatbern Transfers Ten micrograms of total plant DNA was digested with therestrictionenzymes EcoRl, corv, and ffifidhi and etectorphoresed on 0.8% agarose geb for 4 hat 50 V. The DNA within the gel was depurinated with 025 M HO, denatured with 0.4 M NaOH. and transferred by capillary blot technique to Genescreen nylon membranes under alkaline conditions (as described by the manufacturer. DuPont, Boston). Probe DNA (0-2 ug) was labeled with «P by the method of Renberg and Vogebtein 0983) and hybridized to the Southern transfers (Southern 1975). Data Analyses A genotype was assigned to each cultivar for each probe. Because each cultivar represents a single genotype there are no sampling considerations and the gene frequency for a grven auele vrtthin a cultivar is either 0.0, 0.5. or 1.0. Vfe calculated a similarity measure (5) between a pair of cultivars as the proportion of alleles shared in common between cultivars. Suppose for example, that cultivars AandB have genotypes >A and a^a* respectively, and suppose that a, = % a. and a* T 6 a, # a,, then two of the four genes are Identical between A and B, and 5 = 05. Asa second example, suppose the genotypes of A and B are a.a* and a^, respectively. and suppose that a, = a, = a, and a, i* a*a* then 5 = OS. because again two genes are Identical between A and B. if a^az i* a»a, then S OJ0, and if a, = a % = a ora 1 = a, and a* = a«then S = 1-0. The average similarity is XS/n, where mere are n loci. A UPGMA duster analysis was then preformed on the matrix of average simuaribes CSneath and Sokal 1973). Results Polymorphism Among Clones More than 1000 putative clones were screened for inserts of the appropriate size and 20 clones with the appropriat sert size were tested against a panel o ocado DNAs that represented each of three major races. Fifteen of the scree clones were "single copy" in that banding patterns appeared to corresp to those expected from one or two gee loci. Five of the clones produced com banding patterns characteristic of re( ed gene families. An example of the banding patterr sociated with one probe hybridized Southern transfer containing a wide s brum of cultivars is shown in Figure subset of 13 probes that-resolved a 1 ot 14 genetic loci was used to screen s of 36 cultivars and P. duedeana (re seated by the G755A and G755C entr Every cultivar was uniquely identttie the sense that each cultivar was disl from all others for at least one geneti cus (Table 1). The numbers of alleles locus ranged from one (50R1) to si (73RI)- m general there is a high lew polymorphism associated with the pre consistent with the complex breeding tory and hence the broadly based i pool of cultivated avocado. Davis el a! - Genealogical rwabona*ps Among CtMoted Avocado

53 HynCMGnMpl Hybrid Groep 2 ft&mt. Denotogiamol duster iclatlonshps base an a matrix ol the average amber of genes shared I common between pain of cufchars. The ma or cms ters are identified by racial group or as of hybrid a Figaxet. SoattierotnmterolaTOcadoDNA chested with CD«and probed with ckne 64. Cluster Analysis of Genotypic Data Figure 2 presents a dendogram that depicts the relationships among the culbvars screened. Also indicated on the figure are the racial designation or the presumed hybrid origin of the various cultivars. Several features of the data are evident from the dendogram: 0) all cultivars are genotypically unique; (Z) cultivars cluster within racial groupings so the genetic data conform to the botanical designations; (3) many cultivars appear to be of hybrid origin because they are placed in two dusters that fall between the racial groups; and (4) the Guatemalan and West Indian races are more similarto one another than either is to the Mexican race based on RFLP genotype. This pattern is consistent with die presumed geographic origins of the various races (Popenoe W, Zenbneyer G, and Schieber G. unpublished manuscript). Discussion It is important to be able to average over a number of genetic loci because the generic relationships among cultivated avocados reflect a complex breeding history CClegg et al 1993). The transmission path associated with any particular gene may not reflect the average history of the roany loci that comprise the genome. Because we are able to average over a number of polymorphic loci, these dataresolvethe genealogical relationships among avocado cultivars. including patterns of reticulate descent where interracial hybridization was important. The results of this article may be contrasted to those of Mbameed et al 0997). where mintsatewite markers were used to assess relationships among 24 cultivars. Unfortunately many of the cultivars included in the present study differ bom those used by Mbameed et al (1997), so it is not possible to make a comprehensive comparison between the two studies. The cultivars that are common between the two studies do show broadly similar relationships. Thus, for example, the interracial hybrid origin of Hass and Fuetre is also confirmed by Mbameed et al. (1997), however. Hass maps closer to the Mexican representative than to the Giiatwrialanrepreserrtatrtfemtrieu-Rgjirel A knowledge of genealogical relationsnips has already proved useful to the UC Riverside program of avocado improvement. For example, the pollen parent of a sermawarfing cultrvar of considerable commercial promise fjgwen) was unknown. The maternal parent was thecultivar Thille. and it was postulated mat the pollen parent was Nabal; however, the RFLP data exclude Nabal as a pollen parent and suggest instead that Owen resulted from a Thule X Hass cross. As a second example. RFLP genotypes have proved useful in the identification and verification of germplasm materials for the commercial nursery industry. Finally, a rich source of molecular markers is essential for genome mapping and to trace genetic transmission patterns. This latter issue has become particularly important in recent years because there is a crossixmuhanon requirement in avocado that may affect yield. Markers provide an experimental means to establish a relationship between yield and out-pollination by detecting the frequency of outcrossing events among mature fruits. The long generation tunes and complex 322 The Journal of Heredity 199B89(4) UK*V~ TWlt* H^

54 breeding system of avocado has tended to frustrate genetic analysis in the past Molecular markers promise to help overcome some of these obstacles and to provide a wealth of new information. As shown in this study, markers provide a basis for inferring genealogical relationships, thereby revealing new information about the process of domestication. Moreover, molecular markers have proved to be useful tools for asking practical questions to guide avocado management strategies. Undoubtedly a host of additional applications will appear as these techniques become better established. References Bergh BO, ScoraRW, and Storey WB A comparison of the leaf terpines in Penea subgenus fferseo. Bat Gaz 134: Clegg MT, Gaut BS. Duvall MR. and Davis J Inferring plant evolutionary history bom molecular data. H Z J Bot Benberg AP and Vogebtem B A technique lor tadfotabebng DMA restriction endonudase fragments to high specific activity. Anal Bochem 137: Furmer GR. Cimunmgs MR and degg MT, 199ft Evotutioo of the avocados as revealed by DNA restriction fragment variation. J Hered 8UU3-18B. Mbameed S, Sharon D, Kaufman 0, Lahav E, HUM J. DeganiCandUviV Genetic relanonships within avocado (fisrsea mnencana M3iy cnjtivais and between POxa species. Tbeor Appl Genet Popenoe W The avocado: a horticultural pi lem. Trap Agric Rawson JRY. Thomas K. and Clegg MT Puril tion oi tout cellular DNA from a single plant. Biod Genet Sarabtook J. Fritsch EF. and Manlatis T Melee cloning: a laboratory manual. 2nd ed. Cold Spring I bor. New Yorlc Cold Spring Harbor Laboratory Pre SeeyleH CaHomia roots-grower I&28. Sneath PHA and Sokal RR Numerical taxonc San Francisco: W.H. Freeman: Southern EM Detectton of specific sequel among DNA fragments separated by gel wecliuph sis. J Mol Kol Received January Accepted December 13,1997 Corresponding ftiilor. James U Harnrick Oavisetal- Genealogical Relaoorish^ Amor^ Cutavaled Avocado :

55 Appendix III Avocado Rootstock Development by Somatic Hybridization and Genetic Engineering

56 Avocado Rootstock Development by Somatic Hybridization and G Continuing Project; Year 5 of 6 Project Leader: Richard E. Litz (305) ext 310; Department of Horticultural Sciences, TREC; University of Flori Cooperating Personnel: Witjaksono, Simon Raharjo, Miguel Angel Gomez~Lim, Darda Efendi, Isidro Suarez, Pamela Moo Benefit to the industry A new generation of avocadorootstockswith a high level of resistance to Phytophthora root rot (PRR) and with other good horticultural characteristics. Objectives This project has 3 major objectives: I.developing an efficient in vitro protocol for regenerating existing and newly developed rootstocks; 2.genclic transformation of existing avocado rootstock and scion cultivars to enhance resistance to PRR; 3.somatic hybridization of avocado with PRR-resistant Persen species and Nectaiuhitt sp. by means of interspecific protoplast fusion. Summary 1. in Vitro Regeneration improved plant recovery. All of the objectives of this project are dependent upon efficient regeneration of avocado from cell and tissue cultures, i.e., somatic embryogenesis, together with standard micropropagation. During the course of this project, we have optimised these procedures so that they are reliable and adaptable to different genetic engineering approaches (Witjaksono & Litz, 1999a & b; Witjaksono et al,t 1999a; Witjaksono & Litz, 2001a, e), A persistent development problem, however, has been the abnormal development of most somatic embryos failure of the shoot meristem to develop. Consequently, only l%of avocado somatic embryos can form plants. A major accomplishment during the past year has been to increase very significantly the frequency of shoot meristem organization. This has been achieved by pulsing developing somatic embryos With coconut water, and has resulted in plant recovery rates of 10-25%, depending on the cultivar (Witjaksono & Litz, 2001c). Impact; This has had a major impact on genetic transformation and somatic hybridization experiments, and it is now possible to rescue a targe proportion of the genetic diversity that is produced. storage ofembryogenic avocado cultures. Annually, during the course of this project, it has been necessary to re-establish cell and tissue cultures from avocado trees. We have estimated that the time required for this procedure each year is at least one month. During the past year, we have developed a procedure for the long term storage ofembryogenic avocado cultures in liquid nitrogen (Efendi et al., 2001), Embryogemc cultures are immersed in a cryoprotectant solution of DMSO and glycerol for IS min, cooled to -SO^C at a rate of 1 C per min, and plunged into liquid nitrogen ( 196 C). In order to restore growth to cryopreserved cultures, they arc warmed to room temperature, washed thoroughly with avocado plant growth medium and re-plated on plant growth medium (Figs. I & 2). The following cultivars have been successfully cryopreserved: 'Boom 7*. 'Fucrte', 'Gwen\ 'Hass', 'Lula\ 'Reed' and T362\ Impact, Cryopreservatton of avocado cultures will improve the efficiency of our program, ensure continuity of germplasm and ultimately can be used to back up field plantings of avocado genetic resources. 2. Genetic Transformation When plant tissues are infected by plant pathogens, certain proteins are activated in resistant host tissues that restrict the spread of the pathogen. Such proteins have been referred to as pathogcnesis-relatcd (PR) proteins. Many genes that code for PR proteins have been cloned, including chitinase, glucanase, defensin, the antifungal protein, etc. The mycelium walls of Phytophthora species consist of a potyglucan macromolecule, which is degraded by the enzyme glucanase. The procedure for genetically transforming embryogemc avocado cultures wasfirstdefined in our group by Cruz-Hernandez et al. (1998). During the past two years, embryogenic avocado cultures were genetically transformed with constructs that contain a PR, i.e., chitinase, glucanase, chalcone synthase and the antifungal protein (AFP), alone and together with one other PR gene. Because of the limitation imposed by poor somatic embryo development and germination, it has not been possible until the past year to recover large numbers of transgenic plants. Constructs containing the following PR gcne(s) have been used: glucanase, glucanase + AFP, AFP, glucanase + chitinase, chalcone synthase and chitinase. The genes have been cloned in the plasmid pbii21, and have been transferred into avocado using avirulcnt Agrobacieriam tumefaclens strains EHA10I and EHAI05. Embryogenic 'Hass' and 'Fuerte* have have been transformed with several of the constructs, and transgenic somatic embryos are germinating (Figs, 3>7). We anticipate that transgenic avocado plants containing chitinase, AFP and chitinase + glucanase genes will be available for the UC Riverside PRR rootstock evaluation in the next few months. Impact. We anticipate that selections made among the transgenic regenerants wilt have resistance to PRR. 3. Somatic Hybridization High levels of resistance to PRR have been identified in many small-seeded Peneu spp. within the subgenus Eriodaplme (Bergh & Lahav, 1996). The PRR resistant species, including P. horbonia, P. cinerescens and P. pachypoda, are sexually and graft-incompatible with avocado. Somatic hybridization, which involves the fusion of avocado protoplasts with protoplasts of PRR-resistant Persea spp., is one approach for overcoming sexual incompatibility. Although putative somatic hybrids have been recovered, the frequency of successful fusion events is very low, usually less than 0.001%. This, together with the low frequency of normal somatic embryo development noted above, has limited the application of this procedure. In the previous year, a procedure, involving fusion of callus derived protoplasts off. borbonia, P. cinerescens and P. pachypoda with embryogenic avocado protoplasts was adopted. This procedure, used in conjunction with the more efficient method for recovery of normal somatic embryos is now being used on a large scale. Impact. We anticipate that selections made among the somatic hybrids will have resistance to PRR. 4. Other studies Because it is now from somatic emb with 25Gy (LDW) have been recove & Litz, 2001 d). W be available for th months. Impact. Wc anti result in enhanced Eutirm Figure I. Embryoge stained w Figure 2. Recovery Figure 3. Embryoge genes. Bl Figure 4. Hass* som gene. Figure 5. Genetical Figure 6. Germinat Figure 7. Leaf of ge