Comparative Processing and Market Potential of Prune Varieties

Comparative Processing and Market Potential of Prune Varieties Sue Marte NSW Department of Primary Industries (NSW DPI) Project Number: DP04005

DP04005 This report is published by Horticulture Australia Ltd to pass on information concerning horticultural research and development undertaken for the dried prune industry. The research contained in this report was funded by Horticulture Australia Ltd with the financial support of the dried prunes 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 0 7341 1613 6 Published and distributed by: Horticultural Australia Ltd Level 1 50 Carrington Street Sydney NSW 2000 Telephone: (02) 8295 2300 Fax: (02) 8295 2399 E-Mail: horticulture@horticulture.com.au Copyright 2007

COMPARATIVE PROCESSING AND MARKET POTENTIAL OF PRUNE VARIETIES DP04005 (May 2007) FINAL REPORT Susan Marte NSW Department of Primary Industries

DP03001 Susan Marte NSW DPI PO Box 51 YOUNG NSW 2594 p 02 6382 1077 f 02 6382 2228 m 0427 800 379 e susan.marte@dpi.nsw.gov.au June 2007 Photo on front cover: Prune grower field day in Griffith, NSW The purpose of this report is to deliver outcomes from trial work assessing the current Australian prune industry standard prune against prunes used in other world markets to compare their processing and market potential for use under Australian conditions. Funding for this project was provided by the Australian Prune Industry Association, Horticulture Australia Limited and NSW Department of Primary Industries. Any recommendations contained in this publication do not necessarily represent current HAL policy. No person should act on the basis of the contents of this publication, whether as to matters of fact or opinion or other content, without first obtaining specific, independent professional advice in respect of the matters set out in this publication i

LIST OF FIGURES Figure 1. Prune growing regions in Australia 4 Figure 2. Hand harvesting prunes 7 Figure 3. Weighing fruit 7 Figure 4. Mechanical harvest with tarps 7 Figure 5. Wrap-around harvester 7 Figure 6. Loading prunes onto trays 8 Figure 7. Trays stacked on trucks 8 Figure 8. Trucks enter drying tunnel 8 Figure 9. Trucks moving along tunnel 8 Figure 10. Grading prunes into bins 9 Figure 11. Prunes moving through grader 9 Figure 12. Steaming prunes 9 Figure 13. Pitting prunes 9 Figure 14. Grower field day at Griffith 10 Figure 15. Mr Krone pruning van der Merwe 10 Figure 16. Pruned van der Merwe 10 Figure 17. Unpruned van der Merwe 10 Figure 18. Before pruning South Australia site 11 Figure 19. After pruning South Australia site 11 ii

LIST OF TABLES Table 1. Trial data for 2003 Griffith site 12 Table 2. Trial data for 2004 Griffith site 13 Table 3. Trial data for 2005 Griffith site 14 Table 4. Trial data for 2006 Griffith site 15 Table 5. Trial data for 2007 Griffith site 16 Table 6. Trial data for 2006 & 2007 South Australia 18 Table 7. Comparison of fruit attributes over trial years 19 iii

TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES ii iii MEDIA SUMMARY 1 TECHNICAL SUMMARY 2 INTRODUCTION 4 MATERIALS & METHODS 6 Harvest 7 Dehydration & sizing 8 Pruning 10 Fruit assessment 11 RESULTS 12 Griffith 12 South Australia 18 Young 18 DISCUSSION 21 Griffith site 21 South Australia site 22 Water issues 22 TECHNOLOGY TRANSFER 23 RECOMMENDATIONS 24 ACKNOWLEDGEMENTS 25 iv

MEDIA SUMMARY The Australian prune industry standard prune, d Agen, produces a dried product of good size, sugar level and overall quality under optimal Australian growing conditions. Changes to growing conditions, such as limitations of water and high temperatures, challenge Australia s ability to remain competitive in the global market. While representing 2% of prune production on a worldwide scale, Australia under produces for its domestic needs and is reliant on imports from other countries. The ability of the Australian industry to remain price competitive is difficult because of inconsistent fruit quality, high fixed costs and overseas competitors economies of scale. The aim of this project was to benchmark the Australian industry standard prune, d Agen, against the United States standard (California French Improved) and the South African standard (van der Merwe). The industry was looking for a variety which might extend the harvest and processing season. For a variety to be considered suitable it needed to be as good as, or better than, what the industry was already growing and processing. Originally there were three trial sites, two in NSW (Griffith and Young) and one in South Australia. Due to the prolonged drought in Young NSW, that site was removed from the trial. This project was established in 1997. The first four years were spent propagating the varieties to be planted and establishing the trees in the trial sites. Harvesting began at the Griffith trial site in 2003. Assessments of fruit were made during each harvest. When harvested quantities were large enough, fruit was processed according to commercial standards. Harvesting and fruit assessments began at the South Australian site in 2006. While the fruit has been dehydrated in accordance to industry standards, it was not commercially processed due to the small size of the harvest. No fruit was ever harvested from the Young trial site. The South African variety, van der Merwe, is performing well under Australian conditions. Flowering is up to two weeks before d Agen, although fruit maturity has not been that advanced. Propagation and planting of this variety has taken place around prune growing districts. The industry sees promise in this variety. California French Improved, the United States standard, flowers at a similar time to d Agen but early indication shows it maturing up to 10 days prior to d Agen. 1

TECHNICAL SUMMARY The Australian prune industry standard prune, d Agen, produces a dried product of good size, sugar level and overall quality under optimal Australian growing conditions. Changes to growing conditions, such as limitations of water and high temperatures, challenge Australia s ability to remain competitive in the global market. While representing 2% of prune production on a worldwide scale, Australia under produces for its domestic needs and is reliant on imports from other countries. The ability of the Australian industry to remain price competitive is difficult because of inconsistent fruit quality, high fixed costs and overseas competitors economies of scale. In order to remain competitive and to utilise its facilities more efficiently the industry considered ways to extend its growing and harvest season. One way to accomplish this was to plant varieties which would mature before or after d Agen and its clonal varieties. This project was intended to benchmark the Australian standard d Agen prune against that of our competitors the United States standard California French Improved (CFI) and the South African standard van der Merwe. Three trial sites were established. Two in NSW (Young and Griffith) and one in the Riverland, SA. Each prune growing district chose varietal clones of d Agen which they wanted assessed in the trial. The budwood was propagated and grafted, and planted out over subsequent years as material became available by Michael Barkley, NSW DPI. The original trial sites were planted in a randomized complete block design to allow for statistical analysis. Due to co-operator changes and a shift in priorities there was no data gathered for statistical purposes. There were two replicated blocks of 10 trees each at the Griffith site and a planting of 20 trees each of d Agen, van der Merwe and California French Improved at the South Australian site. The trial block at Young was removed due to water issues exacerbated by the sustained drought. The first four years of the trial were spent establishing the trees at the sites. The trees at Griffith had their first harvest in 2003. Assessments made on fruit included: fruit size; yield; soluble solids; fruit firmness; flowering dates; maturity dates; dry-out ratio; and suitability for processing. The information gathered from each harvest was made available to growers at annual field days. Results and progress of the project were reported on at the annual conference of the Australian Prune Industry Association, and information was also included in industry newsletters. Van der Merwe flowers up to two weeks earlier than the industry standard variety, but harvest maturity has been only a couple of days earlier than the standard. under Australian conditions it has cropped heavily, with a large fruit set of generally 2

smaller fruit. There is some indication that larger fruit size can be achieved by adoption of a specific pruning technique, developed by the South African industry. There was interest by growers in the variety which was planted as a buffer row at the Griffith site. Varietal clone B54 showed itself to be early maturing with high sugar content and good fruit size. Assessments of fruit were incorporated into the trial because of the interest it generated. 3

INTRODUCTION There are three main regions in Australia which produce prunes. These are the Murrumbidgee Irrigation Area near Griffith, NSW; the slopes and plains around Young, NSW; and the Riverland area near Renmark, SA (Figure 1). Prune production contributes approximately $25 million to the Australian economy and represents about 2% of worldwide production. Australia under produces for its domestic needs and is therefore reliant on imports from overseas. This is not only costly but exposes the Australian industry to possible pest and disease which may be imported unknowingly. Young, NSW Griffith, NSW Figure 1. Prune growing regions in Australia Renmark, SA The cost to produce prunes in Australia is high. The labour and fixed costs associated with harvest and dehydration are greater than those of our global competitors. To increase our production and to compete in a world market, the industry needs to either produce fruit more efficiently, and therefore less expensively, or extend the harvest season so the cost of the processing equipment can be extended over a longer period. The prune industry is relatively stable compared with other fruit industries. In the history of prune production in Australia, the original variety planted, d Agen, has never been superseded. Although this is positive in terms of knowing that the variety planted will not be obsolete by the time the trees start producing, it also means that the industry is unable to extend its season by planting varieties which bear over time. This ten year project was originated to look at establishing a number of varieties, on a commercial basis, which were identified from the previous HRDC project DAN 7 Prune Variety Evaluation. DAN 7 provided industry with information to select varieties, from a production perspective, that displayed the most promise for processing and consumer acceptance. For any recommendation to be given, though, the suitability of the varieties to processing needed to be assessed in a commercial orchard situation on a long-term basis. Grower groups in the NSW growing regions decided which varieties they wanted to plant for assessment. This also provided an opportunity to import the standard varieties grown from the United States and South 4

Africa, to allow the industry to benchmark the Australian standard d Agen against its competitors California French Improved (USA) and van der Merwe (Republic of South Africa). At the South Australian site, only d Agen, California French Improved and van der Merwe were planted so as to generate data from this prune growing region. Benchmarking varieties is important because it may provide an opportunity for the Australian industry to extend its season of production and utilise its equipment more efficiently. In order for a new variety to be accepted and recommended, it needs to perform as well as or better than, the Australian industry standard. 5

MATERIALS AND METHODS Selected varieties for each district were propagated onto virus tested Myrobalan H29C rootstock by Dr Michael Barkley at the New South Wales Department of Primary Industries Elizabeth Macarthur Agricultural Institute. Planting of nursery trees began during June/July 2000. Because of propagation and quarantine difficulties, the trees were planted out over sequential years as they became available. The trial sites were planted to full capacity by 2002. Varieties planted at Griffith were GF 652, GF 303, GF 698, Cacak Beauty, Cacak Sugar, d Ente 707, California French Improved (CFI), van der Merwe and d Agen. The trial site was planted out as a randomised block design. There were two replications. Each replication had 10 trees of each variety, planted consecutively in rows but the rows were random within the replications. Buffer trees around the perimeter of the block were the d Agen clonal variety B54. Tree spacings were 4 m (between trees) by 6 m (between rows). In Young, the varieties planted were GF 652, GF 303, GF 698, GF 642, d Ente 707, 13/11, California French Improved, van der Merwe and d Agen. At the original trial site 10 trees of each variety (except CFI) were planted. The trial was relocated in 2001 and the rest of the trial planted out. Except for CFI, there were 20 trees of each variety consecutively planted in rows with no replication, in a 4 m x 6 m tree spacing. The propagation of CFI was difficult so the available trees went to the Griffith and South Australian sites. The varieties planted in South Australia were California French Improved, van der Merwe and d Agen. There were 20 trees of each variety planted consecutively in rows, with no replication. Trees were planted 3.5 m between trees and 5 m between rows. To obtain commercially equivalent results, the trees were treated in the same manner as all other trees in the orchard in regards to sprays and pruning. In Griffith, the trial block was adjacent to the commercial orchard, and received similar treatment. This block was originally furrow irrigated but was changed over to a micro-sprinkler system in 2004. In Young, the original co-operator changed production focus and the trial was relocated to a new area of an established orchard within the district. Drip irrigation was installed, but it was not completed until the second year after reestablishment at this site. The trial site in South Australia was incorporated into an existing orchard and was drip irrigated. Originally the project was set up for statistical analysis but after establishment this was no longer an industry priority. The industry felt it was more important the trial be industry acceptable rather than statistically sound. When trees are mechanically harvested the machinery moves through an orchard row by row. It becomes too onerous in a randomised block to record accurate data due to the logistics of mechanical harvest. Observation of the varieties growth habits and processing of the fruit formed the basis of this trial. Soil samples were taken at the Young and Griffith sites before planting. Annual leaf and soil samples were taken at the Griffith site in January of each year since 2002. A 6

pooled soil sample from each replicate was taken and a pooled leaf analysis done for each variety. No samples were taken at the other sites for nutrient analysis. Harvesting Harvest of the Griffith site began in 2003. For the first three years trees were hand harvested, as is industry practice for young trees. Fresh fruit was harvested into picking bags. These were then emptied into fruit lugs (Figure 2). The lugs were individually weighed and totals from each variety recorded (Figure 3). Varieties were kept separated throughout the harvest, drying, sorting and processing stages. Figure 2. Hand harvesting prunes Figure 3. Weighing fruit In 2006 the Griffith trial site was mechanically harvested for the first time. The harvester in 2006 shook the trees, with fruit falling onto tarps which had been placed under the trees (Figure 4). The tarps were then emptied into fruit lugs to be weighed. The lugs were then emptied into wooden fruit bins. Varieties were kept separated throughout the harvest, drying, sorting and processing stages. Figure 4. Mechanical harvest with tarps Figure 5. Wrap-around harvester In 2007, a different mechanical harvested was used (Figure 5). The machine shook the fruit from the trees onto a wrap-around catcher which then emptied onto conveyor belts which emptied the fruit into wooden (or plastic) fruit bins. Fruit which fell below the catcher was manually picked up and emptied into the bin. Varieties were kept separated throughout the harvest, drying, sorting and processing stages. At the Griffith site, trees were harvested by variety within each replicate. The 10 trees were weighed separately and the totals added together which resulted in a final 7

weight for each variety in each of the replicates. A sub-sample of 10 fruit from each variety in each replicate was collected to carry out the laboratory assessments. After the sub-samples were taken and labelled, the fruit from the two replications of the same variety were combined for dehydration and further processing. The dry-out ratio was then tallied for each variety. Harvest began at the South Australian site in 2006. Fruit was hand harvested into lugs in both 2006 and 2007. Each variety was kept separate throughout the entire process. Random sub-samples containing 20 fruit (2 replications x 10 fruit) from each variety were taken for further analysis. Dehydration & sizing The dehydration of the fruit from both trial sites was done by Roy Duffell, a prune grower at Yenda, NSW. The fruit was delivered to the shed in lugs or bins, depending on quantity of fruit, and processed by variety. Each variety was emptied into a water wash which cleaned the fruit and separated any trash (such as leaves) from the fruit. As the fruit came out of the wash it was put onto metal trays with wooden sides (Figure 6). These trays were stacked onto trolleys (Figure 7). Each trolley is referred to as a truck. The trucks were then placed into the parallel flow drying tunnel (Figure 8). The trucks were moved along inside the tunnel according to drying requirements (Figure 9). The tunnels were heated using LPG and were maintained at approximately 80 C. Each truck spent between 16-18 hours in the tunnel, depending on the condition of the fruit. When the trucks exited the drying tunnels they were left to cool. The fruit on the trays was then scraped into wooden bins and stored in this way until needed for processing. Figure 6. Loading prunes onto trays Figure 7. Trays stacked on trucks Figure 8. Trucks enter drying tunnel Figure 9. Trucks moved along tunnel 8

Before processing, the fruit was graded according to size (Figures 10 & 11). From each of the grades a random sample of fruit was taken to obtain the count per pound (count/lb) to make sure the fruit was in the correct grade. Payment to growers is determined by the graded size, with smaller grades (i.e. larger fruit) commanding more money. Figure 10. Grading prunes into bins Figure 11. Prunes moving through grader Fruit was then re-hydrated for further processing. This process can be done in a number of ways. In 2004 the fruit from the trial was re-hydrated using hot baths of prune juice (~ 100 C). The prunes were submerged in the liquid for approximately 5 minutes before being removed, bagged and sealed. In 2006 there was enough fruit to process 10 kg of pitted and unpitted fruit. The fruit to be pitted was put through a steam process in order to heat the fruit through to the pit, which enabled the pit to be removed more easily (Figure 12). A single Ashlock pitter was used (Figure 13). The pitted fruit then went through a sorbate bath before being emptied into bulk cardboard cartons and bagged into plastic Glad Snap Lock re-sealable bags (snack and sandwich size). The unpitted fruit was rehydrated using steam before being mechanically bagged into plastic bags containing 500 g of whole fruit. Figure 12. Steaming prunes Figure 13. Pitting prunes 9

Pruning Minimal tree pruning was undertaken in the first five years after trial establishment. In 2005, one tree from each variety in each replication was detailed pruned at the Griffith site. The trees were pruned using a hand saw, pruning loppers and hand secateurs. During harvest the following year (2006) these trees were harvested separately to compare fruit size and total harvest weight against unpruned trees. In 2006 Mr Chris Krone, a horticultural consultant and chairman of the Dried Fruits Technical Services in South Africa, visited the prune growing regions in Australia to lend technical advice regarding the management of van der Merwe. During the field day in Griffith (Figure 14), he pruned some of the trial trees in Replicate 1 to show specifics to managing the growth habit of this variety (Figure 15). He brought the trees down to 4-5 central leaders and stripped the buds from the top third of each leader. All the van der Merwes in Replicate 2 were pruned in this way during winter 2006 (Figures 16 & 17). These trees were harvested separately from the trees in Replicate 1 to measure any differences between fruit size and characteristics in pruned versus unpruned van der Merwe trees. Figure 14. Grower field day at Griffith Figure 15. Mr Krone pruning van der Merwe Figure 16. Pruned van der Merwe Figure 17. Unpruned van der Merwe 10

The trial site in South Australia was pruned in winter 2006. Most of the trees had strong rootstock growth (Figure 18). This was cut away using hand saws and chainsaws (Figure 19). The tree wounds were not painted or sealed. All trees were pruned back to 4-5 central leaders with the buds removed from the top third of the van der Merwe trees. Figure 18. Before pruning SA site Figure 19. After pruning SA site Fruit assessment A sub-sample of fruit from each variety, at each site, was assessed each year when allowable. The production and quality characteristics of the fruit which were assessed and recorded were: fruit size (g) and (mm); yield (kg); soluble solids ( Brix); fruit firmness (scale of 1-5); colour (scale of 1-5); flowering dates; maturity dates; dry-out ratio; and suitability for processing. The fresh fruit assessments were completed in the laboratory at NSW DPI s Centre for Irrigated Agriculture in Griffith. The samples of 10 fruit were weighed together on an electronic balance. Each fruit was then measured, in millimetres, around its circumference. It was also visually assessed for colour on a scale of 1-5 (1 = green, no red; 2 = ¾ green, ¼ red; 3 = ½ green, ½ red; 4 = ¼ green, ¾ red; 5 = red, no green). Each fruit was then assessed for fruit firmness using a scale of 1 5, with 1 being firm and 5 being soft. This assessment was done by feel only. Fruit was then cut in half to measure the soluble solid content. The juice from the fruit was squeezed onto a refractometer and the Brix recorded. Both hand-held and electronic refractometers were used, depending on what was available at the time. 11

RESULTS Trial sites Griffith - 2003 Harvest began at the Griffith trial site in 2003. All fruit were harvested on the one date which was determined by the sugar reading (Brix ) of the van der Merwe. Although this was not ideal, it did allow for a comparison of maturites between varieties (Table 1). The difference in harvest data between the replications shows differences due to planting years. Most of the varieties in Replicate 1 were planted one year before those in Replicate 2. Van der Merwe was planted in the same year as Replicate 2 and California French Improved in the following year for both replicates. The clonal variety B54 was planted as a buffer row on three sides of the trial site a year after Replicate 1 was planted. This variety has shown to be a very early flowering and maturing variety. Because of interest in these characteristics data from this variety have been incorporated into the results. Table 1. Trial data of fruit assessments for 2003 Griffith site Data summary - 2003 Variety Harvest date Total fresh wt (g) Total no. fruit Ave. fruit wt (g) No. subsample Brix (ave) Ave. size Firmness (mm) Colour 1 2 Cacak Sugar (east) 21/01/2003 85.7 3 28.6 3 15.1 36.5 2 1 698 (east) 21/01/2003 75.3 4 18.8 4 13.9 36.8 2.5 2 d'agen (west) 21/01/2003 61.6 3 20.5 3 16.3 36.7 3.3 2.3 d'agen (east) 21/01/2003 346.4 15 23.1 10 15.5 36.9 3.3 3.9 707 (east) 21/01/2003 2307.7 79 29.2 10 16.2 39.7 2.7 2.5 707 (west) 21/01/2003 1393.7 54 25.8 10 18 39.1 3.5 3.6 303 (west) 21/01/2003 1753.4 66 26.6 10 16 39.6 2.6 3.6 vdm (east) 21/01/2003 344.2 14 24.6 10 19.9 36.5 4.5 4.4 vdm (west) 21/01/2003 5404.8 201 26.9 10 21 39 3.4 3.5 652 (west) 21/01/2003 3962.4 140 28.3 10 17.8 41.6 3.8 3.5 B54 3 21/01/2003 108.1 5 21.6 5 27 33.2 5 4.8 Notes: 1 Colour scale was from 1-5, with 1 being green and 5 fully coloured. 2 Firmness scale was from 1-5, with 1 being very firm and 5 being soft. 3 B54 was planted as a guard row but the fruit was very early and sweet so was also assessed. Griffith - 2004 There was a sizeable crop to harvest in 2004. All trees were in production, except California French Improved. Harvest was planned according to fruit maturity and not date, but the lateness of the season made that difficult. Harvest ended up being over two dates and the resulting Brix levels gave an indication of maturity. There were many more fruit in this season and limited processing was done. This allowed for a dry out ratio and a count/lb (Table 2). The dry out ratio for prunes should be approximately 3:1. This ratio indicates good sized fruit with suitable sugars (Brix 24 and above). The count/lb was an average of six random counts of fruit. The results 12

were in line with expectations, with B54 again being very early to flower and mature. The Cacak series performed poorly when compared with the other varieties and were discontinued from formal assessment. Table 2. Trial data of fruit assessments for 2004 Griffith site Griffith - 2005 In 2005 the crop load was consistent with the age of the trees. Fruit was harvested by hand. Harvest was according to fruit maturity and took place over a number of days (Table 3a). The B54 matured three weeks earlier than the rest of the trial and this made it difficult to get accurate results for processing the fruit as no dehydrator was in operation when that fruit was harvested. These fruit were held in cold storage for 28 days before being dehydrated. Due to poor performance, this was the last year that any fruit was harvested from Cacak Sugar. 13

Table 3a. Trial data of fruit assessments for 2005 Griffith site There was a sufficient quantity of dried fruit in 2005 to allow for mechanical grading. The fruit was passed over a grader (Figure 11) which deposited fruit into bins, according to size (Figure 10). The grades, indicated in Table 3b, are set according to how many pieces of fruit make up one pound (or 454 g). The grade ranges are: <40 (largest fruit); 41 55; 56 70; 71 80; 81 100; 101 110; >110 (smallest fruit). The fruit grade is very important because more money is paid for the larger grades. Also, fruit needs to be a certain grade to pit, and more money is paid for pitted fruit in the retail market. Generally, growers aim to have fruit graded in the 56 70 count/lb or above. Table 3b graphs the grade distribution of the trial fruit. The majority of fruit fell in the 56 70 range, with a good percentage also in the 41 50 range. These grades were very acceptable to the industry. Table 3b. Trial fruit count/lb for 2005 2005 count/lb 100 kgs of fruit graded 80 60 40 20 0 <40 41-55 56-70 71-80 81-100 101-110 vdm CFI d'agen 652 698 303 707 B54 fruit grades 14

Griffith - 2006 In 2006 the trees were mature enough to mechanically harvest. The fruit from each tree was individually weighed and after a sub-sample was taken from each variety in each replicate, the same variety from the two replicates were combined for a total weight. Those trees which had been pruned were kept separate through harvest and dehydration. There were some interesting observations in this year. Prunes are often observed to biennial bear, with a variety bearing a larger crop one year and the next year a lighter crop. This was evidenced by d Agen, 303 and 698 in 2006 (Table 4a). There was also a frost during flowering, when the van der Merwe s were near full flowering and the other varieties only beginning to flower. The frost did not appear to affect the crop on van der Merwe, but may have had some effect on the other varieties and contributed to their reduced yields. Table 4a. Trial data of fruit assessments for 2006 Griffith site As the quantity of fruit harvested increases for a variety, the size of fruit tends to decrease. This does affect prune grading and count/lb. This was clearly indicated with van der Merwe. Also, a larger grade of fruit was seen in the 303 s which had a substantially smaller harvest than the other varieties (Table 4b). Table 4b. Trial fruit count/lb for 2006 2006 count/lb kgs of fruit graded 120 100 80 60 40 20 0 <40 41-55 56-70 71-80 81-100 101-110 vdm CFI d'agen 652 698 303 707 B54 fruit grades 15

The pruned trees were recorded separately to help identify if pruning influenced fruit size. Table 4c tends to suggest that the fruit harvested from pruned trees was larger in weight and diameter, but lower in soluble solids. Table 4c. Effects of fruit size on pruned van der Merwe in 2006 (subsample data) 2006 van der Merwe 45 40 35 30 25 20 Pruned Unpruned 15 10 5 0 mean fruit wt (g) mean fruit size (mm) mean brix attributes Griffith - 2007 Trees were again mechanically harvested in 2007. The pruned and unpruned trees of van der Merwe were harvested and weighed separately. They were also separated for the purposes of analysis, processing and sorting (Tables 7a & 7b) but in the data summary (Table 5a) the analyses from both replicates were averaged. CFI had its first sizable crop in 2007. It matured 7 10 days earlier than the other varieties. Table 5a. Trial data for fruit assessments from 2007 Griffith site The fruit was again mechanically graded according to size. Since there was a large quantity of fruit it was decided to only grade the three comparison varieties d Agen, CFI and van der Merwe (Table 5b). As can be seen from the graph, the majority of the fruit was in the acceptable 56 70 grade range. The rest of the dried fruit was combined after dehydration except 10 kg of each variety which was kept 16

separate. From these 10 kg, three random samples of fruit from each of these varieties were taken to obtain the average count/lb. Almost all the fruit graded into the 56 70 range. The exception was with 698 and 652. These graded into the 71 80 range, which was still acceptable to industry. Table 5b. Trial fruit Count/lb for 2007 2007 count/lb 120 100 kgs of fruit graded 80 60 40 vdm CFI d'agen 20 0 <40 41-55 56-70 71-80 81-100 101-110 fruit grades There was a noticeable difference in the van der Merwe fruit between the pruned and unpruned trees. The fruit from the pruned trees were larger and the sugars were higher than the fruit from the unpruned trees (Table 5c). The fresh weight (Table 7a) from the pruned trees (296 kg) was nearly half that of the unpruned trees (552 kg). The difference in dried fruit weight was not so considerable, with the pruned trees yielding 139 kg and the unpruned trees yielding 172 kg. This meant that the dry out ratio (Table 7b) for the pruned trees (2.1:1) was considerably better than for the unpruned trees (3.5:1). Table 5c. Effects of fruit size on pruned van der Merwe in 2007 2007 van der Merwe 45 40 35 30 25 20 Pruned Unpruned 15 10 5 0 mean fruit wt (g) mean fruit size (mm) mean brix attributes 17

South Australia 2006 & 2007 The trial site in the Riverland, South Australia was under the maintenance of a large commercial grower. There had been maintenance issues associated with the distance of that site from the researcher s base. The trees were to bear fruit in 2005 but due to unseasonably hot weather during blossom there was a crop failure. There were small fruit harvests in 2006 and 2007. Trees were hand harvested in both years. Fruit was harvested into picking buckets and then emptied into fruit lugs. Fruit from each variety was combined to record a total weight. Sub-samples of fruit were then taken for further analysis. The results are summarised in Table 6. Table 6. Trial data for fruit assessments 2006 & 2007 South Australia Young After the original trees in the Young trial site were transplanted to the new location, it was fully planted with all the varieties, except CFI, as they became available over the next couple of years. The transplanted trees survived after being re-established at the new site, but the stress they sustained limited their growth. The newly planted trees died due to a lack of water and they were then replanted. The replanted trees also died due to the continued drought. It was decided by industry in 2003 to discontinue using the Young site in the trial for data collection. The surviving trees have been maintained by the co-operator for use as budwood by a local nursery. No fruit has been harvested from the trees. A comparison of all the fruit attributes for all varieties at the trial sites over the years of the trial are shown in Tables 7a & 7b. 18

Table 7a. Comparison by variety of total fresh weight, total dried weight & average fruit size (mm) from the sub-sample over the trial years 19

Table 7b. Comparison by variety of average fruit weight, dry out ratio & average brix from the sub-sample over the trial years 20

DISCUSSION Griffith site The trial established well at the Griffith site. The project is on target and the results are, for the most part, not surprising. It is unfortunate that the trial site at Young is no longer being assessed and that the site in SA is not doing as well as expected, as it would be beneficial to have the different sites to make a full comparative analysis. D Agen and the clones are producing as expected. There are subtle differences between them and these have been picked up in the assessments. Differences include how they bear fruit (are they biennial, etc) and slight differences in maturity. The one variety which performed unexpectedly was B54. This variety was planted as a buffer row and turned out to be a very early flowering variety and the earliest maturing variety in the block. This variety shows potential as a fresh market fruit although its drying potential is similar to the others. One apparent drawback to this variety is the lack of vigour shown in its growth habit. It has remained a small tree in comparison to the other varieties and has yielded accordingly. The Cacak series (Beauty and Sugar) was disappointing. Tree vigour was lacking in Sugar and the fruit was not satisfactory. Although it showed promise as a recommended variety from the completed HRDC DAN 7 project, it did not perform well in this trial. The fruit was large but very mealy and it did not mature on the tree in accordance to industry needs. The fruit from Beauty was large, flavourful and dried well but the yield from the trees was very small. A possible reason why these varieties did not perform well may be the lack of suitable pollination, as they flower at a different time to the rest of the varieties. Although the trees are still in the trial block, they are no longer being assessed as part of the trial. Van der Merwe has performed well. Although planted a year behind most of the varieties it grew larger and more quickly than d Agen and the clones and cropped more heavily. There are some concerns with it due to the small fruit size associated with heavy cropping. Growers in South Africa find that this variety performs well if it is managed properly. In June 2006 Mr Chris Krone, a horticultural consultant and chairman of the Dried Fruits Technical Services in South Africa, travelled to Australia and visited each prune growing region. He talked about how to manage this variety and gave pruning demonstrations. He said in South Africa they found that by pruning the tree in a certain way they obtained reasonable fruit size. Due to high costs of labour in Australia, prune growers do not prune trees in such a detailed way. This may be the main technical obstacle keeping Australian prune growers from adopting advice to plant this variety. California French Improved has only just begun to bear a sizeable crop of fruit. It was planted two years after d Agen and the clones and one year after van der Merwe. In its initial years of production it followed similar trends to d Agen, in regard to yield. Its growth habit is very upright, with fruit growing inside the tree and up the limbs. In the initial years the fruit matured later than the other varieties. The first sizeable crop was harvested in 2007. The fruit matured earlier than the other varieties and was harvested nine days before d Agen and the clones. In the comparison of 21

attributes it was in line with the other varieties. If this trend continues, with this variety maturing earlier than d Agen yet similar in its attributes, it may have a place in the Australian prune industry as a variety to extend the harvest season. South Australia site Tree loss and trees reverting to rootstock was an issue at this site. Of the 20 trees of each variety planted only 3 van der Merwe, 9 CFI and 7 d Agen are currently productive. Some of the rootstock was regrafted in 2006 with those varieties, to try and increase the number of productive trees. It is too early to comment on the productivity of the comparative varieties at this trial site. The tree habits of the three varieties are more similar to each other at this site than they are in Griffith. This could in part be due to the strong rootstock growth which surrounded most of the trees. Now that the rootstock has been cut away, the trees may fall into their inherent growth pattern. When the rootstock was removed, large areas of cut wood were exposed and this has resulted in some visible infection in those areas. Water issues The viability of these sites, especially the SA site, could be dependent on the current water situation. Both trial sites are irrigated and therefore reliant on availability of irrigation water. The severity of the drought has meant that stringent water reforms have come into place. In SA, at the time of this report, current water allocation is 0%. If there are not sufficient rains this winter (2007) then irrigation next season will be a serious issue and concern. It is understandable if the grower co-operator at this site uses his water for his productive orchard and not the trial site. While there is a chance to replant the site to full capacity, the issue of water availability must be taken into consideration when making this decision. 22

TECHNOLOGY TRANSFER The technology transfer associated with this project has mostly been through workshops and conferences. Annual workshops have been held after each season to provide results from the project to growers in Griffith and Young, NSW. No workshops have been held in South Australia. Each workshop attracted between 10 and 25 people. Reports on this project have been given at the annual conferences of Australian Prune Industry Association in 2005 and 2006. An update on the project is also provided each year for the HAL Annual Industry Report. There was a field day held in June 2006. Mr Chris Krone, a horticultural consultant and chairman of the Dried Fruits Technical Services in South Africa, travelled to Australia to offer advice and share knowledge about the van der Merwe prune variety. He visited Griffith and Young, NSW, to give pruning demonstrations and discuss management issues regarding van der Merwe. The days were well attended with 15 growers in Griffith and 10 growers in Young. While this project helped cover costs of this visit, the visit was initiated and coordinated by the Prune Industry Development Officer, Mr Jeff Granger. 23

RECOMMENDATIONS It is recommended that this trial continues in some capacity. The industry strongly recommends that the van der Merwe and California French Improved varieties continue to be assessed under Australian conditions. While sufficient data are becoming available in regards to van der Merwe, which allows for planting recommendations to be made to industry, this is not yet the case for the CFI. Data from additional seasons needs to be incorporated into current results in order for the industry to feel confident in any recommendation. It is also recommended to continue assessing pruning techniques with the van der Merwe. If the South African pruning technique is too costly under Australian conditions, then pruning to suit both local conditions and the growth habit of this variety needs to be considered. An economic analysis of van der Merwe should be undertaken as to the viability of this variety under Australian conditions. There is some industry support to complete the variety trial planting in South Australia by either re-planting or grafting the missing trees in each variety. There is also the opportunity to plant some recently imported prune varieties for research purposes. The prune Industry Advisory Committee has recommended that this trial becomes a function of the Industry Development Officer position and that maintaining, or adding to, the variety blocks continues on for the foreseeable future. It is a recommendation from the results of this trial that Cacak Sugar and Cacak Beauty be removed from the trial site at Griffith. The trees can either be completely removed or regrafted with a more economically viable variety. An application to continue funding this variety trial has been approved and will commence in July 2007. It is recommended that any remaining funds from this project be carried over into the new project DP07004. 24

ACKNOWLEDGEMENTS There are many organisations and people who have helped to make this project successful by sharing their time and resources. My sincere thanks to Roy Duffell. His tremendous knowledge, input and help in the harvest and dehydration of the trial fruit is very much appreciated. Thanks also to Roy s family for everything they have done, which has helped to make harvest more enjoyable. Thanks to Jeff Granger, Industry Development Manager, who has been a great help in providing information for all things prune. Special thanks to the following for their help at harvesting time: Roy Menzies, NSW DPI; Kathryn Taylor, a work experience student from Sydney University who never knew prunes could be so much fun; Corrie Nieuwkoop and Ken Martin, two of the harvest regulars. Appreciation also to: Bill Berrends, Pikes River; Darren Graetz, grafting & pruning wizard from SARDI; Verity Fruits and Coleambally Irrigation for use of their equipment and resources. 25