Cross-pollination in Olive

Cultivars

by Dr Jenny Guerin and Prof Margaret Sedgley

October 2007

RIRDC Publication No 07/169

RIRDC Project No UA-65A
© 2007 Rural Industries Research and Development Corporation.
All rights reserved.

ISBN 1 74151 564 5

ISSN 1440-6845

Cross-pollination in Olive Cultivars

Publication No. 07/169
Project No. UA65A

The information contained in this publication is intended for general use to assist public knowledge and
discussion and to help improve the development of sustainable regions. You must not rely on any information
contained in this publication without taking specialist advice relevant to your particular circumstances.

While reasonable care has been taken in preparing this publication to ensure that information is true and correct,
the Commonwealth of Australia gives no assurance as to the accuracy of any information in this publication.

The Commonwealth of Australia, the Rural Industries Research and Development Corporation (RIRDC), the
authors or contributors expressly disclaim, to the maximum extent permitted by law, all responsibility and liability
to any person, arising directly or indirectly from any act or omission, or for any consequences of any such act or
omission, made in reliance on the contents of this publication, whether or not caused by any negligence on the
part of the Commonwealth of Australia, RIRDC, the authors or contributors.

The Commonwealth of Australia does not necessarily endorse the views in this publication.

This publication is copyright. Apart from any use as permitted under the Copyright Act 1968, all other rights are
reserved. However, wide dissemination is encouraged. Requests and inquiries concerning reproduction and
rights should be addressed to the RIRDC Publications Manager on phone 02 6271 4165.

Researcher Contact Details

Dr Jenny Guerin Professor Margaret Sedgley
Discipline of Wine and Horticulture Faculty of the Sciences, University of New England
University of Adelaide, Waite Campus Armidale NSW 2351
Glen Osmond SA 5064
Phone: +61 (0)2 6773 2303
Phone: +61 (0)8 8303 6665 Fax: +61 (0)2 6773 3376
Fax: +61 (0)8 8303 7116 Email: [email protected]
Email: [email protected]

In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form.

RIRDC Contact Details

Rural Industries Research and Development Corporation
Level 2, 15 National Circuit
BARTON ACT 2600
PO Box 4776
KINGSTON ACT 2604

Phone: 02 6271 4100

Fax: 02 6271 4199
Email: [email protected].
Web: http://www.rirdc.gov.au

Published in October 2007

Printed on environmentally friendly paper by Canprint

ii
Foreword
The aim of this study was to determine the pollen donors for five major olive cultivars, using paternity
analysis. Bloom time, pollen vitality, and weather conditions during the flowering period were also
recorded. The study was done over two years at two sites and different compatibilities between
cultivars were identified.

The olive industry will benefit from the results of this project through increased productivity. Many
orchards have been established without due regard to the requirements of cross-pollination. These
results will be valuable guidelines to anyone planning new plantings. In addition, they can be used to
increase productivity in established groves through the addition of highly compatible pollen donors.
The effect of introducing compatible pollen donors to a mono-varietal grove could improve yields
from less than 1% up to 4%. New trees can be introduced through grafting onto old trees or inter-
planting. This research also has environmental benefits as a significant increase in yields could be
achieved without expansion of land use, irrigation or addition of fertilisers or chemicals for pest and
disease control.

There were notable similarities between the results obtained at both sites. At least one of the major
pollen donors identified for the mother trees at Gumeracha was also identified as a major donor for
the trees at the NOVA site. This result indicates that these cultivars are effective pollen donors in
more that one environment. The fact that they are major pollen donors at both sites also indicates that
they are highly efficient donors for the respective mother trees.

This project was funded from RIRDC Core Funds which are provided by the Australian Government
and from Olives South Australia, the Queensland Olive Council, the Wylie Group PTY LTD, Michael
Johnston, Hunter Olive Association and Richard and Geraldine Austin.

This report, an addition to RIRDC’s diverse range of over 1600 research publications, forms part of
our New Plant Products R&D program, which aims to facilitate the development of new industries
based on plants or plant products which have commercial potential in Australia.

Most of our publications are available for viewing, downloading or purchasing online through our
website:

• downloads at www.rirdc.gov.au/fullreports/index.html
• purchases at www.rirdc.gov.au/eshop

Peter O’Brien
Managing Director
Rural Industries Research and Development Corporation

iii
Acknowledgements
Olives SA for financial support, and for providing opportunities to speak to industry members at field
days and meetings.
Financial contributors: Olives SA, Queensland Olive Council, Wylie Group PTY LTD, Michael
Johnston, Hunter Olive Association, Richard and Geraldine Austin.
In-kind contributors: Milano Olives, Primary Industries and Resources SA, Silver Acre Olives,
Nangkita Olives, Pinery Olive Plantation.

Milan Pavlic for generously allowing us to sample the trees on his property, Milano Olives, at
Gumeracha SA.
Susan Sweeney for her collaboration over the course of this project with the NOVA collection and for
help with data analysis.
Students: Sonali Mookerjee, for completing the Gumeracha study as part of her Masters degree, parts
of this report have been modified from her thesis entitled ‘Identification of pollen donors for olive
cultivars’. Dr Michele La Mantia and Esmaeil Seify for analysing the Kalamata samples from the
NOVA site. Ji Jing, a visiting scientist from China for extracting DNA from olive embryos.
Jana Kolesik for technical help with field work, pollen vitality assessment and analysis of the Mission
samples from the NOVA site.
Staff from SARDI and AGRF for DNA analysis.
Dr Sophie Gerber, from INRA, France, for responding to all our questions about FAMOZ.

iv
Contents
Foreword...............................................................................................................................................iii
Acknowledgements ..............................................................................................................................iv
Contents ................................................................................................................................................. v
List of Tables .....................................................................................................................................vi
List of Figures ....................................................................................................................................vi
Executive Summary ............................................................................................................................vii
Introduction........................................................................................................................................... 1
Objectives .............................................................................................................................................. 6
Methodology .......................................................................................................................................... 7
Plant material ...................................................................................................................................... 7
Methods ............................................................................................................................................ 10
Data analysis ................................................................................................................................. 12
Results .................................................................................................................................................. 14
Identification of different genotypes in the olive orchards using microsatellite polymorphisms ..... 14
Floral characteristics ......................................................................................................................... 20
Weather data ..................................................................................................................................... 27
Paternity analysis .............................................................................................................................. 27
Flowering survey from different growers properties ........................................................................ 33
Discussion ............................................................................................................................................ 35
Identification of different genotypes................................................................................................. 35
Floral characteristics ......................................................................................................................... 35
Paternity analyses.............................................................................................................................. 36
Conclusions....................................................................................................................................... 38
Implications ......................................................................................................................................... 38
Recommendations ............................................................................................................................... 39
Planting recommendations................................................................................................................ 39
Plantings found to be incompatible under the conditions of this study ............................................ 39
Future research.................................................................................................................................. 39
References............................................................................................................................................ 40

v
List of Tables
1. Olive cultivars used for the study at Gumeracha and the origins of standards used for the
comparison of DNA fingerprints.
2. Genotypes in the NOVA collection.
3. Most likely pollen donors identified for the mother trees analysed at the Gumeracha site.
4. Most likely pollen donors identified for the mother trees analysed at the NOVA site.
5. Property locations and cultivars used in the flowering survey.

List of Figures
1. Field plan of the Gumeracha orchard.
2. Field plan of the NOVA site.
3. Dendrogram of parent cultivars at the Gumeracha site and standard cultivars.
4. Dendrogram of parent cultivars at the NOVA site.
5. Bloom time in olive cultivars at the Gumeracha site.
6. Bloom time olive cultivars at the NOVA site in year 1
7. Bloom time olive cultivars at the NOVA site in year 2
8. Pollen vitality from cultivars in at Gumeracha grove.
9. Pollen vitality in the genotypes at the NOVA grove.
10. The % complete flowers in cultivars at Gumeracha and NOVA.
11. Diagramatic representation of the Gumeracha grove showing the spatial relationship between the
mother trees and their most-likely pollen donors.
12. Flowering times of the cultivars in the flowering survey.

vi
Executive Summary
What the report is about
The aim of this study was to determine the pollen donors for five olive major cultivars, using paternity
analysis. Bloom time, pollen vitality, and weather conditions during the flowering period were also
recorded. The study was done over two years at two sites and different compatibilities between
cultivars were identified.

Who is the report targeted at

The olive industry will benefit from the results of this project through potential increases in
productivity. Many orchards have been established without due regard to the requirements of cross-
pollination. The results from this project will provide valuable guidelines to anyone planning new
plantings. In addition, they can be used to increase productivity in established groves through the
addition of new pollen donors. The effect of introducing compatible pollen donors to a mono-varietal
grove could improve yield (fruit set) from less than 1% up to 4%. New trees can be introduced
through grafting onto old trees or inter-planting. This research also has environmental benefits as a
significant increase in yields could be achieved without expansion of land use, irrigation or addition
of fertilisers or chemicals for pest and disease control.

Background
Olives have been in cultivation for thousands of years, yet pollen compatibility relationships between
cultivars and the mechanism of self-incompatibility have not been fully investigated. In common with
many woody, forest trees, the olive is a wind-pollinated species. Several studies on the identification
of compatible cultivars have led to conflicting results, possibly due to different environmental
conditions under which the studies were conducted, or to confusion in cultivar identity (Mekuria et al.
1999). In addition, self-compatibility and the response to pollen donors can vary between seasons
(Lavee et al. 2002).

The most commonly used methods for identification of compatible cultivars are based on artificial
cross-pollination using pollination bags to regulate the pollen load, and subsequent observation of
pollen tube growth (Wu et al. 2002) and/or fruit set (Lavee et al. 2002). However, conditions inside
the pollination bag may affect fruit set (Rallo et al. 1990) and pollination bags may leak, allowing
pollen from other genotypes to enter (de la Rosa et al. 2004). The results will also vary depending on
whether the initial or final yield is considered because of the high rate of fruit drop in olives (Rallo et
al. 1990).

The identification of the paternal parent using molecular techniques is a reliable method, because the
genetic contribution of alleles is traced from the parents to the offspring. Microsatellite markers are
useful for these studies as they have codominant segregation and a high level of polymorphism in
olive (Sefc et al. 2000; Cipriani et al. 2002; de la Rosa et al. 2002). Several reports have used
microsatellite markers and paternity analysis to study pollen movement in tree populations (Chaix et
al. 2003, Isagi et al. 2004, Robledo-Arnuncio and Gil 2005). The effectiveness of microsatellite
markers in identification of paternal parents of progeny obtained from an olive breeding program has
been demonstrated by de la Rosa et al. (2004).

vii
Results
There were notable similarities between the results obtained at both sites. At least one of the major
likely pollen donors identified for the mother trees at Gumeracha was also identified as a major likely
donor for the trees at the NOVA site. This result indicates that these cultivars may be effective pollen
donors in more that one environment. The fact that they are likely pollen donors at both sites also
indicates that they are potentially efficient donors for the respective mother trees. Selfing was not
significant at the Gumeracha site but was predicted in Frantoio and at a higher rate in Koroneiki at the
NOVA site, probably due to environmental influence.

Implications
To compensate for the variability in pollen viability and flowering times between seasons it is
suggested that more than one pollen donor cultivar is present in a commercial grove, and preferably
three or four. Under the conditions of this study the following combinations of mother trees and
pollen donors were found likely to be compatible:

Frantoio: Kalamata, Mission, Coratina

Kalamata: Frantoio, Koroneiki, Barnea
Koroneiki: Mission, Hojiblanca
Mission: Koroneiki, Arbequina
Barnea: Kalamata, Mission

viii
Introduction
Olives have been in cultivation for thousands of years, originating in the eastern Mediterranean
region, and has since spread to many other parts of the world. The olive tree has played a role in the
commerce and trade of Syria from as early as 3000 B.C. (Connell 1994), and the importance of the
olive tree to the Mediterranean people has been documented by Voyiatzi et al. (1999) and Tsalikidis
et al. (1999). Spain, Italy and Greece are the major producers and consumers of olive products in the
world, although both production and consumption are increasing in areas outside the Mediterranean.
The olive plant was introduced into Australia at the beginning of the nineteenth century (Booth and
Davies 1996), and olives now comprise an important industry in Australia.
The viability of the olive industry depends on harvesting an economic yield of fruit, which can be
used for oil production and table olives. Despite a profusion of flowers on olive trees, only a few set
fruit in a growing season (Cuevas et al. 2001; Ghrisi et al. 1999; Marco et al. 1990; Martin 1990) and
only about 1-2% of these fruits remain on the trees at maturity (Martin 1990). The reasons for this low
fruit to flower ratio include the proportion of male to female flowers, climatic conditions during fruit
set, and compatibility relationships among cultivars (Dal Pero Bertini 1960). Some cultivars are self-
incompatible, which means that the flowers cannot be fertilised by pollen from the same cultivar, and
some cultivars are cross-incompatible, where flowers cannot be fertilised by pollen from certain other
cultivars. Therefore, it is important for growers to understand the cross-compatibility between
cultivars when planning an olive orchard in order to maximise fruit set and yields. This is especially
important when orchards are planted in isolated areas where the only sources of pollen available are
within the orchard.
Since the pollination process may vary from year to year and can be affected by environmental
factors, studying pollen flow amongst different cultivars at the molecular level is a new and exciting
approach that is likely to be a more reliable and powerful method of determining compatibility than
studies based on hand pollination, bagging, pollen tube growth, or fruit set. The identification of
suitable pollen donors using paternity analysis studies analyses the paternal genetic contribution in the
offspring and hence identify the most likely father.

Floral structure and pollen dispersal

Olive trees bear both hermaphrodite and staminate flowers (Ateyyeh et al. 2000) in the form of
panicles (Griggs et al. 1975). The flowers are whitish and small and each bears a ‘short-toothed’ calyx
and ‘short-tubed’ corolla (Hartmann and Opitz 1966). Hermaphrodite flowers generally have two
stamens and a bi-locular ovary with a short style and stigma (Hartmann and Opitz 1966). In staminate
flowers, the pistil is either rudimentary or absent. The flowers are wind pollinated, bear large
quantities of pollen, and no nectar is produced as they do not possess nectaries (Martin 1994). Griggs
et al. (1975) observed that olive flowers are morphologically adapted to either self- or cross-
pollination. In some flowers, the anthers are close enough to the stigma so that when they dehisce, the
pollen falls on the stigma and self-pollination could occur. At the same time there are flowers where
the filaments are flattened so that the anthers spread away from the stigma, thus favouring cross-
pollination. According to Cuevas et al. (2001), olives are naturally suited to cross-pollination by wind.
The presence of flowers with male parts only indicates that these flowers are formed for the sole
purpose of acting as pollen donors. The abundant amount of pollen, up to 200,000 pollen grains per
flower, shows that they are adapted to be wind pollinated.

1
Factors affecting pollination
For successful cross-pollination to occur, it is necessary to have adequate amounts of compatible
pollen available when the flowers are in bloom. This is possible if the compatible cultivars growing in
the orchard have overlapping bloom times. Olive trees of the same cultivar growing under the same
environmental conditions are known to bloom simultaneously (Dal Pero Bertini 1960). Griggs et al.
(1975) and Ghrisi et al. (1999) observed that the bloom dates and duration vary among cultivars and
between years. However, they observed that in most years the bloom time overlapped sufficiently for
adequate pollination. Lavee et al. (2002) studied 36 olive cultivars over a period of 12 years and
observed that the length of the flowering period depended on climatic conditions. Sanz-Cortes et al.
(2002) studied three different cultivars and found that the start of blooming differed for the cultivars
Temprana de Mont, Serrana de Esprad, and Penjoll. These studies highlight the importance of
recording bloom period to ensure the availability of pollen when the stigma are receptive.
The proportion of complete flowers borne by a tree is another factor that has been studied and may
affect the fruit set. Only the complete flowers can bear fruits while the staminate flowers can only act
as pollen donors. Studies by Wu et al. (2002) showed wide variation in the percentage of complete
flowers between cultivars, ranging from 23% in Kalamata to 87% in Picual. Cuevas and Rallo (1990)
observed that trees with a low number of flowers had a higher proportion of complete flowers than
those with more flowers. Rallo et al. (1981) and Lavee et al. (1996) found that although the
proportion of complete flowers differed between cultivars there were no significant differences in the
yield. This may be because of the exceedingly small number of flowers that actually produce fruit.
However, they found that removal of half the inflorescences resulted in double the fruit set in most
cultivars studied. This implies that competition between fruits and non reproductive organs for the
resources of the plant may influence the final fruit set (Rallo et al. 1981). Griggs et al. (1975)
observed that occasionally a tree may bear mostly staminate flowers and thus have a low yield due to
unavailability of enough perfect flowers to mature into fruits. Reale et al. (2006) recently observed
flowers and the rate of aborted flowers at the morphological and cyto-histological level and found that
the number of panicles in the canopy is a major influence on yield rather than the number of flowers
per tree.
Methods of measuring pollen vitality in olives have been studied by Pinney and Polito (1990). They
used both the in vitro germination method and the fluorescein diacetate method to measure the
viability of olive pollen and found that both methods were highly correlated. Similar studies by Wu
(2002) showed that the coefficient of determination between the two methods is 0.86. Pinney and
Polito (1990) also found pollen viability to vary among cultivars. Ascolano was found to have the
highest pollen viability and Mission the lowest. Wu (2002) observed that pollen viability ranged from
a low of 14% in Pendolino to a high of 79% in Frantoio. Occasionally, cultivars have been found to
be male sterile (Moutier 2000; Villemur et al. 1984). It is important to consider the viability of pollen
used in any pollination study since totally male sterile cultivars cannot be pollen donors. Also, there is
no published information on whether the level of pollen viability has any effect on a cultivar being a
pollen donor. Lavee et al. (2002) suggested that since the bloom time and pollen viability of cultivars
vary from year to year, yields may benefit from having than more than one pollen donor in an orchard
in order to ensure adequate pollination each year. Pollen vitality was observed to vary between years
in Arbequina clones (Rovira and Tous 2000). Therefore, multi-year data must be obtained in order to
observe the effect of changes in vitality on the ability of a cultivar to act as a pollen donor.
Several studies have reported that environmental conditions can affect the process of pollination. For
example, Hartmann and Opitz (1966) observed that some cultivars had particular temperature
requirements for optimum pollen tube growth. However, they found that fruit set was not affected by
rain at the time of bloom. Bradley et al. (1961) found that the effect of temperature on growth of
pollen tubes is dependent on the cultivar combination. Low temperatures have been found to reduce
pollen tube growth, and as a result, the pollen tubes are unable to reach the embryo sac before it

2
degenerates (Martin 1994). On the other hand, high temperatures result in faster growth of pollen
tubes (Griggs et al. 1975), although Fernandez-Escobar et al. (1983) and Cuevas et al. (1994)
observed better pollen tube growth at 25°C than at 30-35°C. Androulakis and Loupassaki (1990)
observed that cultivars differ in their compatibility when grown in different environments. They
observed that high temperatures negatively affected pollination. Environmental factors have also been
implicated in fruit set rates (Ghrisi et al. 1999), and the formation of perfect flowers (Lavee et al.
2002). Olives are wind pollinated and therefore the direction of wind during bloom time is an
important consideration.

Problems with fruit set in olives

Several problems may affect the commercial production of fruit. The alternate bearing habit of the
trees causes a heavy crop of small fruits in some years, and a light crop of large fruits in others. Radi
et al. (1990) suggested that the phenomenon is an effect of competition among fruits for the limited
nutritional resources of the tree. A high level of flower and fruit abscission is observed in olives
(Cuevas et al. 2001; Ghrisi et al. 1999; Marco et al. 1990; Martin 1990). As many as 98% of the
flowers have been observed to abscise and only 1 to 2% of the flowers develop into fruits (Martin
1990). However, before abscission, the large number of flowers causes a drain on the resources of the
tree, and this may explain why the highest fruit set rates are followed by the highest fruit drop rates
(Martin 1990). Ateyyeh et al. (2000) reported that there were two stages in perfect flower abscission.
The perfect flowers that were not fertilised abscised three weeks after anthesis, while some of the
fertilised flowers abscised during the next six weeks after anthesis. Olive yields may also be affected
by the formation of shot berries, which are small parthenocarpic fruits of no commercial value
(Rapoport and Rallo 1990). Some researchers have attributed the cause of shot berries to inadequate
cross-pollination (Fernandez-Escobar and Gomez-Valledor 1985; Sibbett et al. 1992).
A major problem faced by both researchers and growers is the confusion regarding the compatibility
relationships for successful fertilisation of olive cultivars. Studies have been conducted in several
different countries to identify the most compatible cultivars (Androulakis and Loupassaki 1990; Rallo
et al. 1990; Antognozzi and Standardi 1978; Sharma et al. 1976; Griggs et al. 1975). While most
studies agree on the benefits of cross-pollination, there is still uncertainty about the most effective
pollen donors for even the most commonly grown cultivars.

Self-incompatibility in olive cultivars

Cuevas and Polito (1997) observed that during self-pollination, most pollen tubes are unable to grow
through the style and reach the ovules for fertilisation, while the pollen tubes arising from cross-
pollination grew faster and reached the ovule. This indicates strongly that a self-incompatibility
system operates in olives. In fact, several studies have shown that some olive cultivars are self-
incompatible. Diaz et al. (2006) studied self-incompatibility by testing seeds obtained from Picual and
Arbequina mother trees using paternity analysis with data obtained from four microsatellite markers.
They found only three incidences of selfing out of the 90 seeds tested and concluded that these two
cultivars were self-incompatible. Wu et al. (2002) tested self-incompatibility by hand pollinating and
observing pollen tube growth and found the cultivars Frantoio, Kalamata, and Verdale to be self-
incompatible. Moutier (2000) studied the compatibility relationships in sixteen olive cultivars and
found that most of these cultivars needed cross-pollination for adequate fruit set. El-Kholy (2001)
reported that the cultivars Pendolino, Leccino, and Coratina are self-incompatible, but he also found
that the compatibility relationships of the cultivars change when grown under different environmental
conditions. This may explain the contrasting results for some cultivars in the literature. The cultivar
Moraiolo was found to be self-incompatible in studies conducted in Italy (Bini and Lensi 1981), while
the same cultivar was found to be self-compatible in studies conducted in India (Singh and Kar 1980).
Leccino was mostly found to be self-incompatible (Antognozzi and Standardi 1978; Ugrinovic and

3
Stampar 1996), although, Bartoloni and Guerriero (1995) found several selections of Leccino to be
self-compatible.
Self-incompatibility could result in low fruit yields in orchards consisting of a single cultivar (Lavee
and Datt 1978; Singh and Kar 1979). Sibbett et al. (1992) observed that cross-pollination could
improve the quality of the crop by reducing the number of shot berries, and a topical application of
Sevillano pollen on Manzanillo reduced the number of small, or parthenocarpic, fruits. Similar
observations were made by Fernandez-Escobar and Gomez-Valledor (1985), and Cuevas and Polito
(1997). Even cultivars that exhibit some level of self fertility have been observed to give higher yields
following cross-pollination (Fontanazza et al. 1980; Lavee et al. 2002).

Cross-compatibility between olive cultivars

Pollen incompatibility has often been found to be the cause of poor performance of olive crops (Ghrisi
et al. 1999; Sibbett et al. 1992). According to Lavee (1998), the pollinating cultivar should be selected
on the basis of economic productivity, quality of pollen grains, compatibility with the receptor
cultivar, and overlapping bloom time. Not all cultivars are cross-compatible, Griggs et al. (1975) and
Cuevas et al. (2001) both found Mission and Manzanillo to be cross incompatible. Wu et al. (2002)
found the cultivars Kalamata, Manzanillo, Pendolino and Picual to be cross-incompatible. Moutier et
al. (2001) reported a high level of cross-incompatibility between French cultivars. They conducted 90
cross-pollinations and found that only 15 cultivar combinations produced good results. Lavee et al.
(2002) also observed cross-incompatibility between cultivars using pollen germination in the presence
of stigma extracts as well as fruit set data.
There is also confusion regarding the identities of some olive cultivars. Inconsistencies in the yield of
cultivars led to DNA fingerprinting studies which have proved that there are problems pertaining to
mistaken identities of cultivars (Guerin et al. 2002; Mekuria et al. 1999; Weisman et al. 1998).
According to Bartolini et al. (1994) there are more than 1200 cultivars throughout the world with
more than 3000 synonyms.
It has been proposed that self-incompatibility in olive is gametophytic (Cuevas and Polito 1997;
Ateyyeh et al. 2000; Wu et al. 2002). In general, reciprocal compatibilities/incompatibilities are
observed in systems under gametophytic control (Lewis 1994; Sedgley 1994). However, other reports
have shown non-reciprocal relationships between olive cultivars (Moutier et al. 2001; Lavee et al.
2002). Lavee et al. (2002) suggested that multiple origins of the domesticated Olea europaea have
resulted in a complex system controlling self-incompatibility. Further work is needed to understand
the genetic system operating in olives. including the identification of the genes involved.

Methods of studying compatibility relationships in olives

Studies on compatibility relationships in olives have so far been based on artificial cross-pollination
and subsequent observation of pollen tube growth (Bartoloni and Guerriero 1995; Cuevas et al. 2001)
and/or fruit set (Singh and Kar 1980). However, when flowers are bagged the microenvironment
differs from the natural conditions under which fruit set usually takes place. In addition, fruit set
observations can be made at different times and may vary considerably as initial fruit set can be high,
but after fruit abscission the final fruit set may not be so impressive. This was particularly evident in
the study conducted by Rallo et al. (1990) where it was observed that although fruit set in self-
pollinated trees was higher at full bloom, after 15 days, open pollinated trees had a higher fruit set at
full bloom after 45 days. Studies in which pollen tube growth is used to determine cross-compatibility
are based on the assumption that if the pollen tube reaches the embryo sac and successful fertilisation
occurs, the fruit will reach maturity. However, it is a common phenomenon that, after the initial fruit
set, a large proportion of the fruits abscise (Cuevas et al. 2001; Ghrisi et al. 1999). Other factors may
be involved during post-fertilisation abscission, which may not necessarily be related to compatibility.

4
De la Rosa et al. (2004) used paternity analysis with data from four microsatellite markers to verify
the parents of seedlings thought to be derived from selfing or controlled crossing. They found a high
level of contaminating pollen had breached the pollination bags during hand pollinations. This finding
throws many previous results using pollination bags into question, especially if the bags were not
placed over the inflorescences well before anthesis. Compatibility relationships based on paternity
analysis trace back the genes in mature fruits from the mother tree and its polliniser, so the true pollen
donor can be identified.

Paternity analysis studies

Several paternity analysis studies have used isozymes as molecular markers to trace the movement of
pollen between cultivars. For example, transmission of genes from parents to offspring in radish was
studied using isozyme analysis where the genotype of the progeny was compared with that of the
mother plant and then the maternal contribution was subtracted to study the paternal contribution of
genes (Ellstrand et al. 1989). A similar approach was used to identify pollen donors for sweet cherry
cultivars Stella and Summit (Brant et al. 1999). It was observed that the presence of a number of
pollen donors in the orchard ensured the availability of adequate pollen at the time of full bloom,
since effective pollination can occur only when the bloom times of the cultivars coincide. Isozymes
were also used to study gene flow in almonds (Jackson and Clarke 1991), where the pollen gene
contribution in the embryos of mature nuts was studied. However, the success of these studies
depended on the availability of isozymes with sufficient polymorphisms to be able to distinguish
between the different cultivars. When few pollen donors are involved, isozymes may prove to be a
useful tool to study gene flow. However, since the polymorphism obtained with isozymes is less than
that required to successfully distinguish between large numbers of cultivars, these markers are not
suitable when many cultivars are present in the population.

Recent studies on paternity analysis have used microsatellite markers that amplify polymorphic repeat
sequence regions of the genome. These markers have been successfully used for paternity analysis in
sweet potato (Buteler et al. 2002) where two microsatellite loci were analysed in the experimental
population. This study also showed that when one of the loci had a lower level of polymorphism in
the population than is required to distinguish between cultivars, it was difficult to analyse the paternal
contribution of genes. Hence, the level of polymorphism exhibited by the marker in the cultivars is an
important factor in these studies. Microsatellite markers have been used confirm Mendelian
segregation using paternity analysis in a population of olives derived from a cross between Leccino
and Dolce Agogia (Rallo et al. 2000). Paternity analysis studies using microsatellite markers have
been used for several tree species including oak (Gerber et al. 2000; Streiff et al. 1999), red pine
(Lian et al. 2001), walnut (Polito et al. 2003) and Eucalyptus (Chaix et al. 2003).

The most likely method was proposed by Thompson and Meagher (1987) and can be used to identify
potential fathers in populations with several genotypes based on the genetic data, since an individual
cannot be a parent unless it has genes in common with the progeny. They use the concept of exclusion
probability, which is the probability that an unrelated individual will be excluded from being a parent.
The likelihood of paternity for each male parent is compared for a particular female parent and her
progeny. The logarithm of likelihood ratios is thus the likelihood of an individual being the parent
compared to all other individuals being the parent. Paternity is assigned to the male with the highest
likelihood value. Thompson and Meagher (1987) used these approaches to identify the most likely
parents in a population of Chamaelirium luteum seedlings based on 11 isozyme markers. The
likelihood based method was also used by Gerber et al. (2000) on a population of oak trees. Gerber et
al. (2000) included estimations of gene flow from outside the stand and gene flow from inside the
stand to minimise errors, and their method has been developed into the computer program, FaMoz,
which has been used successfully to estimate gene flow in a population of Eucalyptus using
microsatellite markers (Chaix et al. 2003). This program was used for our study on olives.

5
Objectives
The aim of this research project is the identification of compatible and competitive pollen donors for
five widely planted olive cultivars in Australia. Compatibility between cultivars will be investigated
using DNA markers and paternity testing. The most likely pollen donors of olive fruits will be
determined by analysing the DNA of embryos collected from trees within the orchard.

6
Methodology

Plant material

Trees
Olive trees growing at Milano Olives, were used for the experiments for the Gumeracha site, South
Australia. Gumeracha is located in the Adelaide Hills at an altitude of 355 m above sea level, latitude
34° 49’ 33” S and longitude 138° 52’ 10” E. The property was laid out in sections supplied by
separate irrigation valves. Some of these sections were planted with a single cultivar, while others
were planted with a mixture of cultivars (Figure 1). The trees in the orchard ranged from one to four-
and-a-half years old at the time of study. The orchard was irrigated with water of 1550 Total
Dissolved Solids (ppm). Tree sprinklers were used at 35 litres per hour and the orchard was irrigated
every two weeks for 10 to 12 hours from December to end of March.

Kalam
ata (1
3) P en
dol
i no
(12
)

UC13A6 (232) 2

Kalamata (215)
1
atypical Kalamata (4)
Pe
ndo
atypic li no
al P e (16
Manz ndoli Kin )
anillo no (1
(22) 8) gK
1 al a

)
ma
(21
ta Ba
(2) rnea
)
(26
(13
io
2 ) 1
Frantoio (69)
n to

2 Ka
i no
ts ou r
Fra

el l a
cc

atypical Frantoio (1)

(5)
Le

S ev
illan
o
(3)
2
2 Verdale (41)
Mission (526)
Koroneiki (249)
atypical Manaki (20)
1

Figure 1. Field plan of the orchard showing the position of cultivars and the selected mother trees
(indicated by 1 and 2). The number of trees for each cultivar are indicated in parentheses. Scale bar =
50m.

7
Fruits collected from mother trees of five cultivars (Barnea, Frantoio, Koroneiki, Kalamata, and
Mission) were used for paternity analysis. Two mother trees of the cultivars were selected, one from a
region of the orchard where it was surrounded by trees of the same cultivar, and one from a location
where it had other cultivars nearby. The reason behind such selection was that the possibility of self-
fertility and the effect of nearby pollen donors could be assessed. Additional trees were used for
collecting fruits when the selected mother trees did not have enough fruits for analysis. All the
genotypes present in the orchard were considered to be potential pollen donors.

Trees at the Roseworthy site (-34 052’S, 138069’E), were planted for the National Olive Variety
Assessment project, established to scientifically evaluate most of the known olive varieties in
Australia, at the University of Adelaide Roseworthy campus (-34 052’S, 138069’E), 50 km north of
Adelaide, South Australia. Roseworthy has a Mediterranean-type climate with an average annual
rainfall of 440mm with 330mm (75%) occurring between the months of April to October. Details of
irrigation and soil conditions are outlined in the RIRDC Publication No 03/054.

The National Collection was planted as a resolvable incomplete block design, consisting of 3
replicates by 2 trees per replicate of 100 accessions sourced from nurseries and old government
collections across Australia. Tree spacing was 6 metres within rows by 7 metres between rows. A
barrier row of olive trees was planted around the 3 replicate blocks. Three mother trees were used,
one from each replicate, so that different neighbours or potential pollen donor varieties surrounded
each one. Figure 2 shows the position of the mother trees relative to the other genotypes planted in the
trial. The mother trees used at this site were Frantoio, Kalamata, Koroneiki and Mission.

8
 7 1 1 2 2 3 3 4 4 29 29 1 1 2 2 3 3 4 4 29 29 7

29 17 17 47 47 23 23 33 33 36 36 48 48 43 43 47 47 1 1 52 52 1

29 24 24 39 39 1 1 22 22 1 1 15 15 42 42 8 8 1 1 14 14 1

4 48 48 15 15 39 39 12 12 54 54 36 36 2 2 15 15 49 49 4 4 2

4 46 46 1 1 10 10 4 4 48 48 39 39 3 3 32 32 19 26 26 2

3 27 27 31 31 1 1 6 6 37 37 33 33 39 39 2 2 22 22 7 7 3

3 5 5 53 53 29 29 46 46 39 39 48 48 13 13 20 20 9 9 30 30 3 R1
2 31 31 44 44 42 42 13 13 3 3 1 1 7 7 51 51 47 47 1 1 4

2 1 1 28 28 18 18 45 45 4 4 1 1 34 34 33 33 50 50 1 1 4

1 40 40 1 1 2 2 41 41 1 1 11 11 13 13 7 7 38 38 29 29 29

1 1 1 46 46 1 1 21 21 29 29 25 25 35 35 45 45 16 16 48 48 29

29 6 6 33 33 4 4 45 45 13 13 41 41 22 22 12 12 21 21 46 46 1

29 2 2 47 47 32 32 33 33 51 51 7 7 8 8 15 15 45 45 20 20 1

4 31 31 47 47 1 1 28 28 44 44 1 1 39 39 15 15 46 46 53 53 2

4 33 33 48 48 39 39 1 1 1 1 11 11 15 15 36 36 25 25 48 48 2

3 39 39 43 43 3 3 34 34 7 7 13 13 42 42 2 2 35 35 13 13 3

3 7 7 52 52 26 26 1 1 1 1 29 29 14 14 4 4 48 30 30 3
R2
2 22 22 1 1 19 50 50 47 47 38 38 1 1 49 49 16 16 9 9 4

2 37 37 36 36 48 48 4 4 3 3 1 1 1 1 54 29 29 39 39 4

1 1 1 23 23 10 10 18 18 42 42 2 2 1 1 39 39 1 1 29 29 29

1 27 27 17 17 46 46 1 1 31 31 40 40 24 24 48 48 1 1 5 5 29

29 27 27 48 48 46 46 4 4 47 47 4 4 33 33 3 3 1 1 1

29 39 39 34 34 1 1 2 2 36 36 48 48 15 15 24 24 39 39 4 4 1

4 1 1 23 23 45 45 7 7 5 5 48 48 1 1 35 35 36 36 45 45 2

4 33 33 3 3 7 7 22 22 14 14 39 39 38 38 44 44 16 16 32 2

3 46 46 1 1 47 47 29 29 31 31 13 13 1 1 1 1 53 53 47 47 3

3 42 42 30 30 13 13 17 17 1 1 49 49 33 33 48 48 1 1 7 7 3
R3
2 15 15 41 41 18 18 28 28 13 13 29 29 20 20 52 52 6 6 9 9 4

2 39 39 25 25 22 22 26 26 21 21 39 39 31 31 19 19 8 8 37 37 4

1 50 50 1 1 46 46 15 15 10 10 1 1 1 1 1 1 1 1 1 1 29

1 2 2 2 2 48 48 42 42 51 51 40 40 43 43 12 12 29 29 11 11 29
7 29 29 4 4 3 3 2 2 1 1 29 29 4 4 3 3 2 2 1 1 7
Figure 2. Field plan of the NOVA site showing the three replicate blocks R1, R2 and R3. Border row
is shaded in light grey. The genotype number of each tree is shown and the mother trees from each
block are indicated by a dark grey square. Mother trees are Frantoio (1), Kalamata (20), Koroneiki
(22) and Mission (42).

9
Flowers
Flowers were collected when the plants were in full bloom. The flowers were collected in plastic bags
and kept on ice during transport from the field to the laboratory. They were then stored at 4°C for a
maximum of 2 days until observations on the percentage of complete flowers were made.

Pollen
Anthers from freshly collected flowers were transferred with forceps to 2 mL Eppendorf tubes, and
left overnight at room temperature in a plastic container containing silica gel. The shed pollen was
used for vitality tests.

Fruits
Fifteen fruits were collected from four sides of each of the selected mother trees and stored in separate
bags. Ten embryos from the fruits in each bag were used for DNA analysis, the extra fruits being
collected to allow for the possibility of parthenocarpic fruits lacking embryos. The fruits were
collected in labelled paper bags and then packed into plastic bags. They were kept on ice while
transporting them from the field to the laboratory and then stored at 4°C until they were used for
DNA extraction.

Methods

Bloom time
The trees were observed at regular intervals of 3-4 days from the commencement of flowering. Start
of bloom was regarded as the time when approximately 10% of the flowers were in bloom, full bloom
when approximately 80% of the flowers were in bloom, and end of bloom when about 80% of the
flowers were spent and the petals had fallen off. The dates for start of bloom, full bloom, and end of
bloom were recorded.

Determination of percentage of complete flowers

Five inflorescences were collected, one from each side of the tree facing north, south, east, and west,
and one from the centre of the tree, and used to study the percentage of complete flowers. Olives have
two types of flowers: complete flowers with well-developed stamens and pistils, and staminate
flowers where pistils are rudimentary or absent. All flowers from each of the five inflorescences were
observed and the number of complete and staminate flowers was recorded.

Pollen vitality tests

Anthers from flowers of five inflorescences of each plant were used for pollen vitality tests. A small
amount of pollen was transferred with a pair of forceps onto a slide and a drop of the vital stain
fluorescein diacetate (2 mg/mL in acetone) was added (Pinney and Polito 1990). A cover glass was
placed on the pollen and the suspension was observed under a Zeiss photomicroscope (Axiophot) at
520nm using an exciter filter of 395 - 440 nm. Pollen grains with active enzymes were observed to
fluoresce whereas non-vital ones did not. The numbers of stained and non-stained pollen grains were
counted in five fields of observation.

Weather data
The weather data were obtained from the Australian Bureau of Meteorology. The nearest weather
station to the Gumeracha olive grove was the Mt. Crawford station. Mt. Crawford is located in the
Adelaide Hills at an altitude of 525 m above sea level, latitude of 34° 43’ 31” S, longitude 138° 55’
40” E and is 11.5 km northeast of Gumeracha. It was assumed that the weather at Gumeracha was
similar to that at Mt. Crawford. The data for the NOVA site was collected from the Roseworthy

10
Agricultural College station at an altitude of 65m above sea level, latitude of 34° 30’ 38” S, longitude
138° 40’ 35” E which is 2 km south of the NOVA site.

DNA extraction from leaves

The method developed by Mekuria et al. (1999) was used to extract DNA from the leaves of the
selected mother trees. About 2 g of leaves were ground in liquid nitrogen to a fine powder using a
chilled mortar and pestle. This leaf tissue was added to 7.5 mL of cold extraction buffer (3% w/v
cetyltrimethyl ammonium bromide (CTAB) (SIGMA), 2 M NaCl, 0.02 M EDTA, pH 8.0, and 1 M
Trizma base) (Steenkamp et al. 1994) and 15 mg of PVP-40T (SIGMA) and 15 μL of 2-
mercaptoethanol (Merck, Damstadt, Germany) were added to the extraction buffer just before use.
The mixture was thoroughly shaken and placed on ice. The samples were incubated at 60°C for 30
min with inversion every 10 min and then cooled on ice. After addition of 7.5 mL of chloroform-
isoamyl alcohol (24:1), the tubes were spun on a rotary wheel for 10 min at approximately 60 rpm
followed by centrifugation in a swing-out head centrifuge at 4,000 rpm. The upper aqueous layer was
transferred to a clean tube and two-thirds volume of cold isopropanol added. The tubes were covered
with Parafilm® and inverted gently. The DNA was spooled onto a glass rod and left in 20 mL of wash
buffer (76% (v/v) ethanol and 10 mM ammonium acetate) until it was white. Then the DNA was
dissolved in 1mL of TE buffer (10 mM Tris HCl, 0.1 mM EDTA, pH 8.0). RNA was removed by the
addition of 2 μL of 100 μg/mL DNase-free RNaseA (AMRESCO®, Solon, Ohio, U.S.A.,) and
incubation for 30 min at 37°C. The solution was transferred to a sterile tube and 2 mL of TE buffer
and 1 mL of 7.5 M ammonium acetate was added and mixed. The tubes were kept on ice for 20 min
to allow the proteins to precipitate, and then centrifuged at 10,000 rpm for 20 min at 4°C. The
supernatant was transferred to another tube and 2 volumes of cold ethanol were added. The tubes
were again kept on ice for 20 min for the DNA to precipitate, and then centrifuged at 8,000 rpm for 10
min at 4°C. The supernatant was removed and the excess ethanol was allowed to evaporate before the
DNA was dissolved in 1 mL of TE buffer and stored at –20°C till required.

The DNA quality was checked by observing the absorbance in a UV-160A, UV-visible
spectrophotometer (Shimadzu Corporation, Tokyo, Japan). The ratios of the absorbances at 260/280
and 260/230 were calculated. DNA having ratios of 1.8 or above was considered to be of good quality
(Ausubel et al. 1987) and were used for further analysis. The DNA concentration was calculated from
the absorbance at 260 nm, considering that 1 O.D. corresponds to 50 μg/ml of double-stranded DNA.

DNA extraction from fruits

Since DNA was to be extracted from a large number of embryos, a quicker method of DNA extraction
than the one used for leaves was required. A DNA extraction method modified from the protocol
obtained from the web page of Tree Genetic Engineering Research Cooperative
(www.fsl.orst.edu/tgerc/dnaext.htm) was used to extract DNA from embryos. The flesh was sliced
from the olive fruits and the stone cracked open with a vice. The embryo was separated from the
endosperm using a pair of forceps and immersed in 500 μL of grinding buffer (100 mM Tris, pH 8.0,
20 mM EDTA, pH 8.0, with 4 mg/mL diethyl dithiocarbamic acid sodium salt and 100 μg/mL DNase
free RNase (AMRESCO®, Solon, Ohio, U.S.A.) added just before use). The embryo was ground in
the buffer using a micropestle (Eppendorf) attached to a drill and kept on ice until all the samples
were ready. Then the samples were incubated for 10 min at 65°C followed by the addition of 500 μL
of lysis buffer (100 mM Tris, pH 8.0, 20 mM EDTA, pH 8.0, with 1 M NaCl, 2% SDS, and 1%
sodium metabisulphite added just before use) and incubated for another 30 min at 65°C. The tubes
were inverted every 10 min during this time to mix the contents. The tubes were cooled on ice and
then 1 mL of phenol:chloroform:isoamyl alcohol (25:24:1) was added. The tubes were spun on a
rotary wheel for 10 min at approximately 60 rpm, centrifuged for 10 min at 14,000 rpm and the
supernatant removed to a fresh 1.5 mL tube. 500 μL of isopropanol was added to each tube and kept
on ice for 15 min. The tubes were centrifuged for 5 min at 14,000 rpm and the supernatant removed.

11
The pellets were washed with 1 mL wash buffer (76% ethanol and 10 mM ammonium acetate) on a
rotary wheel for 10 min at approximately 60 rpm. The tubes were spun at 13,500 rpm and the
supernatant was then decanted and the DNA pellet dried under reduced pressure in a Labconco
Centrivac Concentrator at 45°C for 10-15 min. The pellet was suspended in 50 μL of TE buffer (10
mM Tris HCl, 0.1 mM EDTA, pH 8.0). The DNA quality and concentration were determined as
described in Section 3.2.5.

PCR
DNA amplification was performed in a volume of 25 μL containing 60 ng genomic DNA, 1.5 mM
MgCl2, 1 x PCR Buffer (200 mM Tris-HCl (pH 8.4), 500 mM KCl), 0.2 μM of each dNTP, 0.4 mM
of forward and reverse primers, and 1.25 units of Taq DNA Polymerase (Invitrogen). A negative
control was also added which contained the reagents of the PCR mix but did not contain the DNA.

A Programmable Thermal Controller (M.J. Research Inc. USA) was used for amplification of the
DNA. The DNA amplification procedure consisted of the following steps: initial denaturation at 95°C
for 5 min, 34 cycles of 45 s at 95°C, 45 s at the appropriate temperature for the primer, 45 s at 72°C,
and finally an extension step at 72°C for 45 min.

Primers and their annealing temperatures were selected from published articles: SSR3, SSR4, and
SSR14 (Sefc et al. 2000) UDO6, UDO8, UDO24, UDO31 (Cipriani et al. 2002) EMO 2 (de la Rosa
et al. 2002) were used to type the parents and progeny from the Gumeracha grove. The primer
UDO31 was replaced by SSR9 (Sefc et al. 2000) for the NOVA population as the rate of PCR failure
with UDO31 was high. The forward primers were labelled with fluorescent phosphoramidites (FAM,
NED, or HEX), which are detected at wavelengths of 518 nm, 580 nm and 556 nm to enable scoring
after passing through the ABI Prism 3700 DNA Analyser. An internal size standard (ROX), which is
detected at the wavelength of 605 nm was included in each run.

ABI Prism 3700 DNA Analyser

For the Gumeracha data the PCR products were analysed using a 96-capillary ABI Prism 3700 DNA
Analyser (Applied Biosystems). FAM and HEX labelled primers were obtained from GeneWorks Pty.
Ltd., and the NED labelled primers were obtained from Applied Biosystems. The data were analysed
using the program Genescan version 3.5.1 (Applied Biosystems) and the alleles for each locus were
scored by comparing with the bands obtained with ROX which gave peaks at known intervals of
molecular weight. All PCR amplifications and capillary electrophoresis separation were done in
duplicate. A third replicate was done only if consistent results were not obtained with the first two
experiments.

The NOVA data were analysed on an ABI 3770 DNA Capilliary sequencer using Genotyper software
(Applied Biosystems). The PCR and data analysis were done by the Australian Genome Research
Facility (AGRF) at the Waite Campus, SA. PCR results that were ambiguous were repeated.

Data analysis

Identification of parental genotypes

The presence or absence of alleles was recorded in a Microsoft Excel spreadsheet. The Numerical
Taxonomy Systems Software Package, NTSYS-pc version 2.02k (Rohlf, 1993) was used to compare
the fingerprints of the potential parents by preparing the data in the form of dendrograms. A similarity
matrix was prepared using SIMQUAL (Similarity of Qualitative Data) and a dendrogram was
constructed using the UPGMA (Unweighted Pair Group Mathematical Average) and SAHN
(Sequential, Agglomerative, Hierarchial, and Nested Clustering Methods) algorithm (Rohlf, 1993).

12
Paternity Analysis
Paternity analysis (Gerber et al. 2003) was generated using the software program FAMOZ
(www.pierroton.inra.fr/genetics/labo/Software/Famoz/index.html). The statistical analysis used in this
program is based on the most likely method described by Meagher and Thompson (1986). A
simulation was done using 10000 offspring to obtain the threshold for paternity in order to minimize
errors due to gene flow from inside and outside the stand. Distributions of LOD scores from the
simulated population were plotted against the offspring generated from the genotyped parents. The
crossover point between the two graphs was taken as the threshold as described by Chaix et al.
(2003). The genotype with the highest log of the odds ratio (LOD) score was considered as the most
likely father. The LOD score represents the likelihood of a particular genotype being the father
compared to all other genotypes. Famoz settings error was set at 1% which is the error likely to be
generated with microsatellite markers (Oreilly et al. 1998).

Estimating the probability of null alleles present in the data set.

When only one allele was present it could not be determined whether a homozygote or heterozygote
with a null allele was measured, so the second allele was scored as missing data to avoid incorrect
exclusion of potential fathers. This rule was used for embryo data from the Gumeracha property.

For data from the NOVA site we estimated the probability of null alleles according to (Khadari et al.
2003) and three of the primers (SSR3, SSR9 and SSR14) were found to be unlikely to have null
alleles at the 1% level. In these cases where only one allele was detected we assumed homozygosity,
and using this method more pollen donors were identified above the threshold level.

Identification of genotypes in the orchards

Genetic similarities among all pairs of individuals were estimated using the simple matching
coefficient. Cluster analysis was performed on the estimated similarities using the unweighted pair
group method with arithmetic average (UPGMA) and the SAHN algorithm, and the resulting clusters
were expressed as a dendrogram using NTSYS-pc (Exeter Software v.1.8) (Rohlf, 1993).

13
Results

Identification of different genotypes in the olive orchards using

microsatellite polymorphisms

1. Gumeracha
DNA that was extracted from the leaves of the parent trees was compared to DNA from standard
cultivars from the DNA database using the DNA fingerprinting technique described in the methods
section. Table 1 shows the list of standards used for the cultivars in the grove. The data obtained from
the DNA analyses of the selected trees showed that the fingerprints of most of the mother trees tested
matched their corresponding standard DNA samples. However, a few trees that did not match the
standards were found. The cultivar referred to as Manaiki by the grower did not match with the
standard Manaiki in the database and was referred to as atypical Manaiki (a-Mk). The Pendolino
plants in the orchard were sourced from two different nurseries. The Pendolino obtained from one
nursery matched the standard Pendolino, but that sourced from the other did not and was referred to as
atypical Pendolino (a-Pen). One Frantoio plant that was morphologically different to the other
Frantoio trees showed a different fingerprint compared to the standard Frantoio DNA and was
referred to as atypical Frantoio (a-Fra). Some trees that were identified by the grower had a different
appearance from the other Kalamata trees. DNA fingerprinting confirmed that these trees were a
different genotype and they were named atypical Kalamata (a-Kal).

A dendrogram showing the relationship of the tested trees with the standards is shown in Figure 3. It
can be seen on the dendrogram that all the cultivars matched the standards except for the atypical
genotypes mentioned above. Seventeen different genotypes were identified in the grove using the
DNA marker set.

14
Table 1. Olive cultivars used for paternity analysis study at Milano Olives, Gumeracha, South
Australia, and the origins of the standards used for the comparison of DNA fingerprints.

Cultivar Source of the standard

Barnea The Volcani Centre, Bet-Dagan, Israel
Frantoio Consiglio Nationale delle Ricerche, Instituto di Ricerca Sulla Olivicultura,
Perugia, Italy.
Kalamata Consiglio Nationale delle Ricerche, Instituto di Ricerca Sulla Olivicultura,
Perugia, Italy.
Katsourella NOVA Collection, Roseworthy Campus, University of Adelaide, South
Australia, Australia.
Koroneiki Subtropical Plants and Olive Trees, Institute of Chania Agrokipio, Chania,
Greece.
King Kalamata NOVA Collection, Roseworthy Campus, University of Adelaide, South
Australia, Australia.
Leccino The Volcani Centre, Bet-Dagan, Israel
Manzanillo The Olive World Collection, Centro de Investigacion y Desarollo Agrario,
Cordoba, Spain.
Manaiki NOVA Collection, Roseworthy Campus, University of Adelaide, South
Australia, Australia.
Mission Foundation Plant Material Service, University of California, Davis,
California, United States.
Pendolino Consiglio Nationale delle Ricerche, Instituto di Ricerca Sulla Olivicultura,
Perugia, Italy.
Sevillano The Olive World Collection, Centro de Investigacion y Desarollo Agrario,
Cordoba, Spain.
UC13A6 NOVA Collection, Roseworthy Campus, University of Adelaide, South
Australia, Australia.
Verdale Foundation Plant Material Service, University of California, Davis,
California, United States.

15
 B ar
B ar 1
B ar 2
a-Kal
Ma
a- Fra
Ka t
Ka t 1
Ka t 2
a-M a 1
a-M a 2
M is
M is 1
M is 2
M is 3
M is 4
M is 5
Sev
Sev 1
Sev 2
Sev 3
a-Pen
UC
UC 1
UC 2
UC 3
Kal
Kal 1
Kal 2
Kal 3
Kal 4
Pen
Pen 1
Pen 2
Pen 3
KKal
KKal 1
KKal 2
Man
Man 1
Man 2
V er
V er 1
V er 2
Ko r
Ko r 1
Ko r 2
Fra
Fra 1
Fra 2
Le c
Le c 1

0.50 0.63 0.75 0.88 1.00

Geneti c D i stance

Figure 3. Dendrogram showing comparison of the cultivars present at Milano Olives, Gumeracha, SA
with the standard cultivars (Table 1).
Bar: standard Barnea; Bar1 and Bar2: Barnea trees; Fra: standard Frantoio; Fra1, Fra2: Frantoio trees;
a-Fra: atypical Frantoio; Kal: standard Kalamata; Kal1, Kal2, Kal3, Kal4: Kalamata trees; a-Kal:
atypical Kalamata; Kat: standard Katsourella; Kat1 and Kat2: Katsourella trees; Kor: standard
Koroneiki; Kor1, Kor2: Koroneiki trees; KKal: standard King Kalamata; KKal1, KKal2: King
Kalamata trees; Lec: standard Leccino; Lec1: Leccino tree; Man: standard Manzanillo; Man1, Man2:
Manzanillo trees; Ma: standard Manaiki; a-Ma1 and a-Ma2: atypical Manaiki trees; Mis: standard
Mission; Mis1, Mis2, Mis3, Mis4, Mis5: Mission trees; Pen: standard Pendolino; Pen1, Pen2, Pen3:
Pendolino trees; a-Pen: atypical Pendolino; Sev: standard Sevillano; Sev1, Sev2, Sev3: Sevillano
trees; UC: standard UC13A6; UC1, UC2, UC3: UC13A6 trees; Ver: standard Verdale; Ver1, Ver2:
Verdale trees.

16
2. NOVA
Results from DNA fingerprinting the trees in the NOVA collection are shown as a dendrogram in
Figure 4. The results are very similar to those previously obtained with RAPD markers (Guerin et al.
2002) and several differently named cultivars are shown to have the same genotype. Only one
difference was detected with the microsatellite markers; the Mission accession #10 was slightly
separated from the Frantoio group. A total of 54 individual genotypes were distinguished with the set
of primers used. The genotype numbers of the cultivars planted at NOVA are shown in Table 2. Each
genotype was considered as a potential pollen donor in the paternity analysis.

17
 Frantoio 1
Emu Flat 13
Paragon 26
Frantoio 95
Mediterranean 29
Belle de Espagne 40
Boothby ’s Lucca 46
Pueblana 86
Bouteillon 49
Corregiola 52
Palsano 80
Frantoja 57
Leccure 66
Lucca 68
Corregiola 72
Mission 10
Buchine 50
Queen of Spain 27
Areccuzo 35
Verdale 93
Arbequina 5
Big Spanish 41
Dr Fiasci 55
Black Italian 42
FS17 58
Wallace 28
Koraneiki 32
Picual 2
Picual 12
Azapa 22
Hojiblanca 8
Oje Blanco Doncel 77
Del Morocco 54
O de Grasse 75
Nevadillo Blanco 74
Ascolano 36
Attica 39
Mission 71
177 11
Columella 51
Gros Reddeneau 60
Tarascoa 91
Marchiosa 69
Salome 90
Hardy’s Mammoth 61
Katsourella 31
Pigale 83
Rouget 88
Amelon 34
Barouni 15
Regalaise de Languedoc 87
Nab Tamri 18
Sevillano 19
Cucco 53
Macrocarpa 70
Polymorpha 84
Pendulina 80
Volos 94
Borregiola 47
Picholine 82
Longue d ’Ascoli 67
Morihoso 73
Gaeta 59
Frantago 56
Oblonga 78
Souri 33
Coratina 9
Atro Rubens 37
Manzanillo 4
Manzanillo 16
Oblitza 76
Blanquette Early 44
Leccino 6
Pendolino 7
Jumbo Kalamata 25
Atroviolacea Brun Riber 38
Institute 62
Large Fruited 63
Large Fruiting 64
Palermo 79
Praecox 85
Mission 17
Verdale (SA) 21
Black Italian 24
Benito 23
Blanquette Late 45
Verdale (NSW) 92
Bouquettier 48
Large Pickling 65
Blanquette 43
Rubra Baillon d ’Aise 89
Kalamata 30
Barnea 3
UC13A6 20
Manaki 14
0.50 0.63 0.75 0.88 1.00
Genetic Distance

Figure 4. Dendrogram derived from the data obtained with 8 microsatellite markers for each of the
cultivars planted at the NOVA site, NOVA accession numbers are shown in parenthesis. Groups of
identical genotypes have a genetic similarity of 1.

18
Table 2. Genotypes in the NOVA collection. The table shows the given genotype numbers and the
cultivar names we have used for that genotype number in this report. The NOVA numbers of the
cultivars that have the same genotype and the names of the synonymous cultivars in the collection are
also shown.
Genotype # Cultivar Name NOVA #'s Synonyms from NOVA
1 Frantoio 1, 13, 26, 29, Corregiola, Paragon, Emu Flat,
40, 46, 49, 52, Mediterranean, Belle de Espagne,
57, 66, 68, 72, Boothby's Lucca, Pueblana,
80, 86, 95 Bouteillon, Palsano, Frantoja,
Leccure, Lucca.
2 Picual 2, 12, 96
3 Barnea 3, 97
4 Manzanillo 4, 16, 98
5 Leccino 6
6 Pendolino 7
7 Hojiblanca 8, 100
8 Coratina 9
9 Mission (WA) 10
10 177 11
11 Manaiki 14
12 Barouni 15
13 Sevillano 18, 19, 53. Nab Tamri, Cucco.
14 UC13A6 20
15 Verdale (SA) 17, 21, 24 Mission, Black Italian.
16 Azapa 22
17 Benito 23
18 Jumbo Kalamata 25
19 Queen of Spain 27
20 Kalamata 30
21 Katsourela 31
22 Koraneiki 28, 32 Wallace
23 Souri 33
24 Amelon 34
25 Areccuzo 35
26 Ascolano 36
27 Atro Rubens 37
28 Atroviolacea Brun Ri38
29 Arbequina 5, 41, 99 Big Spanish
30 Black Italian (BEO) 42
31 Blanquette 43, 89 Rubra Baillon d'Aise
32 Blanquette Early 44
33 Bouquettier 45, 48, 92 Verdale (NSW), Blanquette Late
34 Buchine 50
35 Columella 51
36 Del Morocco 54, 75 O' de Grasse
37 Dr Fiasci 55
38 FS17 58
39 Hardy's Mammoth 60, 61, 69, 90, Gros Reddeneau, Marchiosa,
91 Salome, Tarascoa.
40 Institute 62
41 Large Pickling 65
42 Mission 39, 71 Attica
43 Nevadillo Blanco 74
44 Oblitza 76
45 Oblonga 56, 78 Frantago
46 Palermo 63, 64, 79 Large Fruited, Large Fruiting
47 Pendulina 81
48 Picholine 47, 59, 67, 73, Borregiola, Gaeta, Morihoso,
82 Longue d'Ascoli.
49 Pigale 83
50 Praecox 85
51 Regalise de Languedo87
52 Rouget 88
53 Verdale (BEO) 93
54 Volos 94

19
A note about Frantoio and Verdale genotypes.
Many differently named cultivars had the same genetic fingerprint as Frantoio with the markers used
in this study as well as the RAPD markers used in a previous study analysing the genotypes planted in
the NOVA site. The cultivar planted at the Gumeracha grove which was named Corregiola had the
same fingerprint as the Frantoio standard from Italy. We have referred to it as Frantoio for the
purposes of this report.

In contrast three cultivars named Verdale were planted at the NOVA site and all had different
genotypes, known as homonyms. Verdale with the NOVA number 21 is commonly known as SA
Verdale and had the same genotype as the Verdale at Gumeracha and the Verdale sample obtained
from The University of California. We have referred to it as Verdale for the purposes of this report.

Floral characteristics
Events at bloom time contribute significantly towards fruit yield in olive. Studies by Cuevas and
Polito (1997), Fernandez-Escobar and Gomez-Valledor (1985) and Wu et al. (2002) showed that
some cultivars may be self-incompatible and others cross-incompatible, and concluded that olive
orchards should contain several cultivars in order to maximise the possibility that all cultivars will
receive compatible pollen. This raises the question of identifying cultivars that are not only cross-
compatible, but also have overlapping bloom times, to ensure adequate pollen transfer at the time the
stigmas are receptive. The ultimate fruit yield in any given season depends to a large extent on the
processes that occur during the flowering period. We measured bloom time and pollen vitality for the
cultivars grown at the two sites as well as sex ratio for the mother trees and some other commonly
grown cultivars, over the two years of study.

Bloom time and pollen viability

As cross-pollination is most likely to occur when compatible cultivars flower simultaneously, the
bloom periods for each genotype were recorded in the two years of the study (Figures 5-7). The dates
for the beginning of flowering, full bloom and the end of flowering are shown in Figure 5 for the
Gumeracha property and Figures 6 and 7 for the NOVA site. The maximum temperature and rainfall
are also shown on the graphs.

20
 Year 1 November December
Date 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3 4 5 6 7 8 9 10 11 12 13
T (¡C) 22 29 32 26 23 16 24 25 32 24 20 16 18 24 21 19 27 25 17 19 18 16 22 20 14 16 #
R (mm) 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.2 4.0 19 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.4 4.0 0.0 0.0 0.0 0.2 Trees
Cultivar
Bar (2) 2 2
a-Fra (1) 1 1
Fra (3) 1 2 90
Kal (5) 5 228
Kor (4) 1 3 249
Kat (2) 2 5
Kkal (3) 1 1 1 26
Lec (4) 3 1 11
a-Mk (4) 2 1 1 17
Man (3) 2 1 22
Mis (5) 1 1 1 2 526
Pen (3) 1 2 16
a-Pen (1) 1 18
Sev (1) 1 3
UC (5) 1 1 3 232
Ver (4) 1 3 42

Year 2 November December

Date 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3 4 5 6 7 8 9 10 11 12 13
T (¡C) 23 25 27 27 26 19 19 26 27 19 16 20 27 30 33 19 19 17 19 #
R (mm) 0.2 0.0 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.4 1.0 0.0 0.0 0.0 0.0 0.9 0.8 0.0 1.0 Trees
Cultivar
Bar (2) 1 1 2
a-Fra (1) 1 1
Fra (2) 1 1 90
Kal (5) 5 228
a-Kal (1) 1 4
Kor (4) 4 249
Kat (2) 2 5
Kkal (3) 3 26
Lec (4) 4 11
a-Mk (4) 1 3 17
Man (3) 3 22
Mis (5) 2 3 526
Pen (3) 1 2 16
a-Pen (1) 1 18
Sev (1) 1 3
UC (5) 4 1 232
Ver (4) 3 1 42

Figure 5. Bloom Time in Olive Cultivars at Gumeracha, during the two years of study. The bars
indicate the time from start (10% of flowers open) until the end (80% spent) of flowering. T:
Temperature (°C) and R: Rainfall (mm) for each day are shown. Cultivar names are abbreviated as
follows: Bar, Barnea; a-Fra, atypical Frantoio; Fra, Frantoio; a-Kal, atypical Kalamata; Kal,
Kalamata; Kor, Koroneiki; Kat, Katsourella; KKal, King Kalamata; Lec, Leccino; a-Mk, atypical
Manaiki; Man, Manzanillo; Mis, Mission; Pen, Pendolino; a-Pen, atypical Pendolino; Sev, Sevillano;
UC, UC13A6; Ver, Verdale trees. The number of trees tested from each cultivar are shown in
parentheses, and of those, the number of trees reaching full bloom on each day are shown on the
graph. Total number of trees for each cultivar in the grove are shown in the final column.

21
 Year 1 October November
Date 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 Total
Max temp 22 26 31 19 22 20 20 23 27 34 22 23 32 26 21 19 25 33 37 29 28 31 41 24 trees
Rain (mm) 0.0 0.0 0.0 3.2 0.6 1.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 in bloom
Cultivar #
1 36 1 6 25 68
2 13 2 15
3 3 8 1 12
4 9 2 1 4 1 17
5 1 2 1 4
6 5 1 6
7 12 1 2 2 17
8 5 5
9 1 3 2 6
10 2 3 5
11 2 2 4
12 2 4 6
13 16 2 18
14 1 5 6
15 15 3 18
16 2 2 2 6
17 1 4 1 6
18 2 1 3 6
19 1 1 2
20 2 2 2 6
21 Few Flowers 1
22 2 1 9 12
23 3 1 1 5
24 1 0
25 1 2 1 1 5
26 1 1 1 3 6
27 1 4 5
28 3 3
29 11 3 4 18
30 2 2 1 1 6
31 11 1 1 1 1 15
32 1 1 2
33 1 1 6 1 1 2 12
34 No Flowers 0
35 2 4 2
36 2 6 2 10
37 1 4 5
38 1 1 4 6
39 6 2 16 1 1 2 1 29
40 No Flowers 0
41 6 6
42 6 1 5 12
43 1 3 1 1 6
44 5 1 0
45 6 6
46 4 3 9 16
47 8 1 8 17
48 1 7 2 17 1 28
49 2 4 6
50 Few Flowers 1
51 1 2 1 1 5
52 2 4 6
53 No Flowers 0
54 2 2 1 5

Figure 6. Bloom Time in Olive Cultivars at the NOVA site in year 1. The bars indicate the time from
start (10% of flowers open) until the end (80% spent) of flowering. T: Temperature (°C) and R:
Rainfall (mm) for each day are shown. The genotype number for each cultivar is shown in the first
column. All of the trees at the site were tested and the number of trees reaching full bloom on each
day are shown on the graph. Total number of trees for each cultivar that bloomed in the grove are
shown in the final column.

22
 Year 2 November
Date 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Total
Max temp 17 20 23 30 28 27 26 27 26 31 30 26 34 39 27 31 36 41 30 25 20 21 24 27 32 36 35 33 trees
rain (mm) 1.6 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.2 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 in bloom
Cultivar #
1 2 8 9 33 17 7 4 2 82
2 2 2 10 1 15
3 1 9 2 12
4 7 3 6 1 17
5 3 1 4
6 4 2 6
7 4 11 15
8 2 2
9 2 3 1 6
10 2 3 1 6
11 2 1 1 2 6
12 3 2 1 6
13 1 1 5 6 5 18
14 1 4 1 6
15 2 1 3 12 18
16 2 3 1 6
17 2 3 1 6
18 1 1 1 1 2 6
19 1 1 2
20 3 2 1 6
21 2 1 1 1 5
22 11 1 12
23 3 1 1 5
24 3 2 5
25 3 3 6
26 6 6
27 2 1 2 5
28 3 2 5
29 1 3 1 2 9 2 18
30 2 2 2 6
31 2 6 3 1 12
32 1 1
33 17 1 18
34 2 1 1 4
35 1 1 4 6
36 1 1 6 3 1 12
37 2 1 3
38 1 5 6
39 2 8 16 2 1 29
40 1 1 1 3 6
41 2 4 6
42 2 5 2 1 1 11
43 2 1 1 2 6
44 2 1 3 6
45 2 2 2 6
46 1 9 6 16
47 6 4 5 2 1 18
48 1 4 1 14 1 3 2 2 28
49 3 2 1 6
50 4 2 6
51 5 1 6
52 1 2 1 2 6
53 1 1 3 1 6
54 2 1 3

Figure 7. Bloom Time in Olive Cultivars at the NOVA site in year 2. The bars indicate the time from
start (10% of flowers open) until the end (80% spent) of flowering. T: Temperature (°C) and R:
Rainfall (mm) for each day are shown. The genotype number for each cultivar is shown in the first
column. All of the trees at the site were tested and the number of trees reaching full bloom on each
day are shown on the graph. Total number of trees for each cultivar that bloomed in the grove are
shown in the final column.

The cooler climate in the Adelaide Hills had the effect of a delayed bloom time by approximately 24
days in year 1 and 17 days in year 2 at Gumeracha, compared to the NOVA site in the northern plains
at Roseworthy. During the first year several trees did not bloom in the NOVA collection as they were
still overcoming juvenility. All cultivars flowered in the second year. At Gumeracha, Frantoio,
Verdale and atypical Frantoio had fewer flowers than the other cultivars as they had only been planted
for one year compared to the other trees that were between 2 and 4.5 years old.

23
Bloom times of the potential pollen donors overlapped to some extent with the maternal trees during
both years of the study. The only exception was in year 1 when atypical Frantoio did not begin to
flower until the flowers of Barnea, Frantoio, Koroneiki, and Mission were already predominantly
spent. The atypical Kalamata trees were not identified until the second year and, therefore bloom
times were not recorded for these trees in the first year of study.

Pollen vitality was measured to confirm that the trees produced viable pollen for fertilization. At
Gumeracha, percent pollen vitality ranged from 24% in King Kalamata to 72% in Koroneiki in the
first year and from 20% in UC13A6 to 66% in Leccino in the second, as shown in Figure 8. Very little
pollen was collected from Verdale and atypical Manaiki in both years, indicating that, at least for the
trees sampled, these two cultivars were predominantly male sterile. Pollen vitality varied between
years and was generally higher in the first year of study.

Figure 8. The % vitality of pollen from the cultivars in the grove at Gumeracha tested in A) year1 and
B) year 2.

At the NOVA site pollen vitality ranged from 11% for UC13A6 to 71% for Arecuzo in the first year
and from 8% for Regalaise de Langedoc to 78% for Mission in the second year, shown in Figure 9.

24
Verdale was also found to be predominantly male sterile at the NOVA site in both years. No general
trend was observed at this site regarding increased pollen vitality in either year.

Figure 9. The % vitality of pollen from the genotypes at the NOVA site tested in A) year1 and B) year
2.

25
Floral structure
The inflorescence of the olive flower is a panicle with flowers distributed along the rachilla, or main
axis, in an opposite pattern. Olive flowers can be complete, having both pistil and anthers or
staminate, where they lack a functional pistil and cannot set fruit. The number of complete flowers per
inflorescence is expressed as % complete flowers and data from both study sites are shown in Figure
10. The mean % complete flowers per panicle at the Gumeracha site ranged from 42 to 98 and
differed significantly between years for 9 out of the 16 cultivars tested. At the NOVA site the means
ranged from 14 to 100% and there were marked decreases in the number of complete flowers in the
cultivars Pendolino, Manzanillo and Mission in the second year. The low levels seen at this site would
still be capable of producing a good yield for olives, approximately 2 - 4% fruit set. However if the
number of flowers capable of setting fruit were below about 10% it may result in low yields.

Figure 10. The % complete flowers from five panicles tested over two years from selected cultivars at
A) Gumeracha and B) NOVA.

26
Weather data
Olives are anemophilous, requiring dry conditions and adequate air movement for pollen distribution.
The weather during the bloom periods was recorded from the Bureau of Meteorology. The weather
station used for the Gumeracha property was at Mt Crawford, which is located at a distance of 11.5
km north-east of the study site. The bloom periods were mostly dry, with rainfall occurring on only a
few days. The maximum temperatures during the bloom period of both the years ranged from 15°C to
33°C with an average of 21°C. However, it was between 20°C to 30°C on most days, which is
suitable for pollen tube growth (Cuevas et al.1994). Maximum temperatures and mm rainfall are
shown in Figure 5. Winds were predominantly easterly and westerly in the mornings and mainly
easterly in the afternoons, with winds of lower intensities blowing in other directions through out the
day.

The NOVA site data were recorded from the weather station at Roseworthy Agricultural College.
Only a small amount of rain was recorded during bloom time at this site over both years. Maximum
temperatures were warmer on the northern plains and ranged from 17°C to 41°C with an average of
28°C. On several days the maximum reached higher than 30°C in both years which may have effected
pollen tube growth (Figures 6 & 7). Wind movement was predominantly north-easterly in the
mornings, shifting to south-westerly in the afternoons.

In general, conditions were suitable for effective pollen dispersal in both of the orchards during the
flowering periods.

Paternity analysis

Gumeracha
The pollen donors that were assigned to the embryos sampled from the maternal trees are shown in
Table 3. The LOD scores for the most-likely fathers ranged from 0.10 to 10.88, with an average of
3.63, and were above the estimated threshold for paternity, which ranged from 0.01 to 0.06, with an
average of 0.02. The numbers of embryos assigned to pollen donors ranged from 169 for Koroneiki,
represented by 249 trees, to one embryo for atypical Frantoio represented by only one tree in the
grove. Paternity could not be assigned to approximately 22% of embryos; these are listed as having an
unknown pollen donor. Unknown donors were most commonly found for Mission embryos, with only
57% having been assigned a pollen donor from the cultivars in the grove.

27
 Table 3. Numbers of embryos assigned to each of the likely pollen donors over two years.
Cultivar names were abbreviated as follows: Bar, Barnea; Fra, Frantoio; Kor, Koroneiki; Kal, Kalamata; Mis, Mission; Kat, Katsourella; KK, King
Kalamata; Lec, Leccino; Pen, Pendolino; Man, Manzanillo; Sev, Sevillano; UC, UC13A6; Ver, Verdale; a-F, atypical Frantoio; a-K, atypical
Kalamata; aMk, atypical Manaiki; a-P, atypical Pendolino; Ukn, unknown.

Pollen donor: Bar Fra Kor Kal Mis Kat KKal Lec Pen a-Mk Man Sev UC13A6 Ver a-F a-Kal a-Pen Unknown

Mother Tree
Barnea 0 4 0 18 30 3 8 3 29 0 0 15 9 13 0 3 2 23

Frantoio 1 0 0 29 45 1 0 2 6 27 2 6 0 3 0 0 2 36

Koroneiki 3 0 0 12 68 4 1 3 8 8 1 2 18 2 0 2 2 26

Kalamata 0 26 87 0 3 0 2 2 1 8 0 2 3 1 1 0 2 22
28

Mission 1 6 82 0 2 1 0 0 0 0 0 0 0 0 0 0 0 68

Total 5 36 169 59 148 9 11 10 44 43 3 25 30 19 1 5 8 175

 The major difference between the two years of the study was due to the activity of UC13A6,
which was estimated to pollinise one embryo in year 1 compared to 29 in the second year. UC13A6
had the third largest number of trees after Mission and Koroneiki, yet its overall contribution to the
embryos was relatively low (4%) compared to Mission (19%) and Koroneiki (21%).
Kalamata was a major pollen donor for Barnea in both the years. However, Pendolino and
Sevillano were also among the top pollen donors in the first year, while Mission, Verdale, UC13A6,
and King Kalamata were more significant pollinisers in the second year. The main pollen donors
identified for Frantoio were Mission, Kalamata, and atypical Manaiki. Kalamata was predominantly
pollinised by Koroneiki and Frantoio. Mission and Koroneiki had a clearly reciprocal relationship, as
each was the major pollen donor for the other. The only incidence of self-pollination detected
occurred in two embryos from Mission.
The spatial relationship between the mother tree and the most likely pollen donors, and the
frequencies of pollen donation are represented in Figure 11. In some cases it appears that
incompatibilities exist between the maternal trees and potential pollen donors. For example, pollen
from Koroneiki fertilised many flowers from Kalamata and Mission, but did not contribute to the
embryos tested for Barnea or Frantoio. Similarly, Kalamata was assigned as the likely pollen donor
for many embryos from Koroneiki, Frantoio, and Barnea, but not for those from Mission. UC13A6
was a significant pollen donor for Koronieki, but made a weak contribution to Kalamata and Barnea,
and was not represented among the donors for the embryos from either Frantoio or Mission. The
results of this study have been published previously (Mookerjee et al. 2005).

29
 Kal Pen Kal Pen

UC Kal UC Kal

a Kal a Kal

Lec

Lec
a Pen Pen ÆBar Æ Kat a Pen Pen Bar Kat
Man KKal Sev Man KKal Sev
a/Cor Ver Æ a/Cor Æ Ver

Mis Kor Mis Kor

a Kor a Kor

Barnea Corregiola

Kal Pen Kal Pen

UC Kal UC Kal Æ
a Kal a Kal
Æ
Lec

Lec
a Pen Pen Bar Kat a Pen Pen Bar Kat
Man KKal Sev Man KKal Sev
a/Cor Ver a/Cor Ver
Æ
Mis Kor Mis Kor

Æ a Kor a Kor

Koroneiki Kalamata

Kal Pen
Key to embryo numbers
UC Kal 0
1-10
a Kal
11-20
Lec

21-30
a Pen Pen Bar Kat 31-40
Man KKal Sev 41-60
a/Cor Ver >60
Æ
Mis Kor
a Kor
Æ
Mission
Figure 11. Diagrammatic representation of the olive grove showing the spatial relationship between the mother trees studied (Æ) and the
most likely pollen donors. Shaded areas indicate the number of embryos fathered by that cultivar ranging from none (white) to >60 (black)
as indicated by the key. Cultivar names were abbreviated as follows: a/Fra, atypical Frantoio a Kal, atypical Kalamata; a Man, atypical
Manaiki; a Pen, atypical Pendolino; Bar, Barnea; Kal, Kalamata; Kat, Katsourella; KKal, King Kalamata; Kor, Koroneiki; Man, Manzanillo;
Mis, Mission; Pen, Pendolino; Sev, Sevillano; UC, UC13A6; Ver, Verdale.

30
NOVA
Likely pollen donors that were assigned to the embryos sampled from the maternal trees are shown in
Table 4. The LOD scores ranged from 1.10 to 6.16, however the threshold was determined at 2.2 and
only scores above this were recorded as indicating the true parent (Gerber et al. 2000). A total of 586
likely pollen donors were identified out of 920 embryos analysed. The most embryos assigned to
pollen donors were from Koroneiki (45), Hardy’s Mammoth (25), Manaiki (23), Coratina (21),
Mission WA (21) and Kalamata (20). The cultivars with the least number of embryos identified were
Bouquettier (0), 177 (1), Verdale (1), and Oblitza (1).

31
 Table 4. Most likely pollen donors identified for the mother trees analysed at the NOVA site. The cultivar name, replicate # and year are shown for
the mother trees in the left-hand column. The genotype number of the likely pollen donors are shown in the top row. The number of embryos
estimated to be pollinised by each genotype are indicated in the table. The total number of embryos attributed to pollen from each genotype are shown
in bold for each mother cultivar and the grand total over the whole study. For some of the embryos the likely pollen donor could not be distinguished
between two cultivars and these have been placed under the genotypes 1/9, 2/16 and 4/44.

genotype # 1/9 1 2/16 2 3 4/44 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 total

Frantoio 1 yr 1 4 1 3 1 1 1 2 1 6 1 1 1 1 1 2 2 1 2 32
Frantoio 1 yr 2 2 2 1 1 1 1 2 1 1 1 1 1 1 1 2 1 1 21
Frantoio 2 yr 1 2 2 1 1 1 2 1 1 2 1 1 1 1 1 1 5 1 1 1 1 1 29
Frantoio 2 yr 2 2 1 1 1 2 3 1 1 2 1 1 2 2 2 1 23
Frantoio 3 yr 1 1 3 1 2 1 3 1 2 1 1 3 1 1 1 2 1 1 1 27
Frantoio 3 yr 2 1 1 5 4 1 3 4 1 1 1 1 1 1 1 26
Total 2 10 15 9 0 6 0 0 3 2 9 6 0 10 5 1 2 0 4 3 1 3 13 0 2 0 3 4 1 0 0 1 0 0 1 0 2 2 0 1 1 13 0 0 6 0 0 3 2 5 1 0 2 0 2 2 0 158
Koroneiki 1 yr 1 1 1 2 1 1 3 2 1 1 5 1 1 2 2 1 1 1 1 28
Koroneiki 1 yr 2 2 1 1 1 1 1 1 1 1 1 1 3 1 1 1 1 1 20
Koroneiki 2 yr 1 1 1 5 1 1 2 1 1 3 1 1 1 2 1 2 2 1 1 28
Koroneiki 2 yr 2 1 3 2 1 2 1 1 1 3 2 1 2 1 2 1 2 1 27
Koroneiki 3 yr 1 5 3 1 2 1 2 1 2 1 2 1 1 22
32

Koroneiki 3 yr 2 3 2 1 2 1 1 4 2 1 1 1 1 1 21
Total 0 1 13 5 0 5 0 0 2 9 7 2 0 3 1 0 0 0 5 6 3 4 4 1 16 1 3 9 1 2 0 0 0 1 0 0 2 0 2 1 0 8 0 1 9 1 0 0 4 5 3 0 1 2 1 2 0 146
Kalamata 1 yr 1 1 1 2 1 2 1 1 1 7 1 1 2 2 1 3 1 1 29
Kalamata 1 yr 2 3 2 1 2 1 2 1 1 2 2 1 1 1 1 1 1 1 24
Kalamata 2 yr 1 2 1 1 2 5 1 1 1 1 1 1 1 1 1 3 23
Kalamata 2 yr 2 4 1 2 1 1 2 1 2 1 1 3 2 1 1 1 1 1 1 27
Kalamata 3 yr 2 3 2 1 4 2 2 1 1 1 1 1 2 1 1 23
Total 10 0 1 0 8 0 0 2 2 2 3 10 0 5 0 2 0 0 1 3 4 2 2 7 16 0 2 1 0 1 0 0 1 1 3 0 1 3 1 1 2 1 0 4 0 5 1 2 0 1 3 4 2 2 0 1 3 126
Mission 1 yr 1 1 1 1 1 3 3 3 3 1 1 1 2 1 1 1 24
Mission 1 yr 2 1 1 1 1 3 1 2 1 4 2 4 1 1 2 1 26
Mission 2 yr 1 1 1 1 4 1 1 2 1 1 1 1 2 2 2 21
Mission 2 yr 2 1 2 1 2 1 1 1 2 3 3 1 1 2 1 22
Mission 3 yr 1 1 1 1 2 2 1 1 3 4 1 2 1 3 1 2 1 1 1 1 1 2 1 34
Mission 3 yr 2 5 1 2 1 3 3 2 2 6 1 2 1 29
Total 0 0 2 1 7 1 0 1 2 4 2 3 0 5 3 0 6 1 1 0 4 4 1 1 11 2 11 2 3 1 3 10 3 12 2 0 1 1 3 1 1 3 1 8 4 0 0 9 0 1 0 5 1 1 1 2 4 156
Grand Total 12 11 31 15 15 12 0 3 9 17 21 21 0 23 9 3 8 1 11 12 12 13 20 9 45 3 19 16 5 4 3 11 4 14 6 0 6 6 6 4 4 25 1 13 19 6 1 14 6 12 7 9 6 5 4 7 7 586
The total number of Frantoio embryos where likely pollen donors were identified was 158. The
predominant likely donors were Kalamata, Hardy’s Mammoth, Coratina and Manaiki. There were
also some selfed embryos identified in Frantoio. Selfing also occurred in Koroneiki with 16 embryos
identified as being selfed progeny. Likely pollen donors were identified for 146 embryos analysed
from Koroneiki. Other prominent donors estimated for Koroneiki were Mission, Hardy's Mammoth,
Hoji Blanca and Arecuzo. Kalamata had likely pollinisers identified for 126 of the embryos analysed,
they were mainly attributed to Koroneiki, Mission WA and Barnea. Frantoio was also a potential
pollen donor as pollinisers for 10 embryos were either from Mission WA or Frantoio. Mission had
likely pollen donors identified for 156 embryos. The major donors were Blanquette, Koroneiki,
Amelon, Arbequina and Oblonga.

The cultivars Frantoio and Mission WA appear to be closely related as their genotypes were not
distinguished using RAPD markers and only differed by one microsatellite peak in this anlaysis. They
appear close together on the dendrogram, Figure 4, and were difficult to distinguish from each other
as pollen donors. In several cases they received the same LOD score so the true parent could not be
identified. Similarly the cultivars Picual and Azapa, as well as Manzanillo and Oblitza received the
same LOD score for some of the embryos. In these cases scores were added to a separate column
where either cultivar could have donated pollen to the embryo: Frantoi/Mission WA (1/9),
Picual/Azapa (2/16), and Manzanillo/Oblitza (4/44). In this way these potential parents were not
eliminated from the data set.

Flowering survey from different growers properties

Several growers were asked to mark down the flowering times for three olive trees from each of the
main cultivars grown on their properties. The property locations and cultivars are listed in Table 5
below.

Table 5. Property locations and cultivars used in the flowering survey

Property Location Cultivars

Keith SA Frantoio, Barnea, Leccino, Picual, Manzanillo
Willunga SA Frantoio, Corregiola, Manzanillo, Kalamata, Mission, Koroneiki
Nangkita SA Frantoio, Barnea, Leccino, Kalamata, Koroneiki, Pendolino, UC13A6
Muswellbrook. NSW Corregiola, Hardy’s Mammouth, Manzanillo
Gunnedah NSW Barnea, Manzanillo, Hardy’s Mammouth, Kalamata, Jumbo Kalamata,
Mission

The results from the growers are shown in Figure 12, and clear differences in flowering times were
observed between the different environments. However, within each property bloom times for the
cultivars overlapped such that it would have been possible for any cultivar to be a pollen donor for the
others providing that they were cross compatible. However, full bloom, when the maximum amount
of pollen is released, did not always coincide for all of the cultivars.

Frantoio or Corregiola was present in the SA properties and the NSW properties. Flowers from these
trees first opened in Muswellbrook approximately one month before the trees at Keith. Trees at
Willunga and Nangkita reached full bloom simultaneously about a week later than the Keith property.

33
Site/Cultivar October November December
Keith SA
Frantoio
Barnea
Leccino
Picual
Manzanillo
Willunga SA
Frantoio
Corregiola
Manzanillo
Mission
Koroneiki
Kalamata
Nangkita SA
Frantoio
Barnea
Leccino
Koroneiki
Kalamata
Pendolino
UC13A6
Muswllbk. NSW
Corregiola
Manzanillo
Hardy's Mammoth
Gunnedah QLD
Barnea
Manzanillo
Kalamata
Hardy's Mammoth
Mission
Jumbo Kalamata

Figure 12. Results from the flowering survey. Data supplied by participating growers for 3 trees of
each cultivar. Bands show the bloom time for each cultivar from the start of flowering (10% of
flowers open), full bloom (darker coloured squares) to the end of flowering (80% of flowers spent).

34
Discussion
The results presented in this report demonstrate that microsatellite markers and paternity analysis are
effective tools to survey the frequency of pollen donors for selected maternal trees in a mixed orchard.
This report documents the first studies using these techniques to monitor pollination patterns in an
olive grove. By using paternity analysis to study the distribution of pollen donors within the grove the
actual events of fertilization under the conditions of free, open pollination have been observed. The
results from this study show that at the time of stigma receptivity in the maternal tree, viable pollen
from the donor was present in the air, could travel to the recipient flower and successfully compete
with other pollen grains on the stigma to reach the egg cell. The successful pollen donors identified in
this report are therefore not only compatible with the maternal tree but have been demonstrated to
function effectively under field conditions.

Identification of different genotypes

The identification of the different cultivars in the groves was done primarily to determine the
genotypes present that could act as pollen donors. The marker set was adequate for distinguishing all
the genotypes in the stands. In the case of the Gumeracha property the DNA fingerprinting found that
some a-typical trees had been planted as different cultivars. The identification of atypical genotypes
was useful information for the grower, and also ensured that these genotypes could be included as
potential pollen donors. Such a situation may often be encountered in an olive orchard since there is a
lot of confusion about the genetic identity of cultivars.

The cultivars planted in the NOVA collection had previously been DNA typed using 10 RAPD
markers (Guerin et al. 2002), and were typed again with microsatellite markers. The two sets of
markers gave remarkably similar results, which shows that the RAPD analyses were robust. However,
it was necessary to type the collection again with the microsatellites for the paternity study so the
parents and embryos could be analysed with the same marker set. The codominant segregation of
microsatellites allows for more efficient parentage analyses compared with dominant markers (Gerber
et al. 2000) such as RAPDs.

Floral characteristics

Bloom time
Recording the bloom times was important to determine that all the potential pollen donors were
shedding pollen at the same time as the maternal trees were in flower. However, the results also
showed that some of the other cultivars were not synchronized in their bloom times. In the first year at
the NOVA site, Dr Fiasci flowered late and did not overlap with the earlier flowering cultivars
Barouni, and Oblonga. In year 2 Coratina had a very short flowering time, which finished at the time
that Manaiki was just beginning to flower. These results show that not all cultivars will bloom at the
same time and certain cultivars may not be suitable as pollinisers for some trees purely on the basis of
bloom time synchronicities.

Pollen viability
The lack of pollen grains in the anthers taken from Verdale indicated that the trees tested were male
sterile in both years from both sites. However, paternity analysis showed that Verdale contributed
pollen to 19 embryos at Gumeracha but to only one embryo from the NOVA samples. Pollen vitality
observations were taken on only a sample of flowers collected at full bloom stage. It is possible that
some flower clusters that were not sampled had viable pollen, and also that viable pollen developed

35
later in the season after the trees had been sampled. The high incidence of male sterility in this
Verdale genotype means indicates that it would not be recommended as a pollen donor cultivar in a
commercial olive grove.

The cultivars UC13A6 also had low pollen vitality in both sites in one of the years and King Kalamata
had low levels measured in both years at Gumeracha. These cultivars were not major pollen donors at
either of the sites, which may be due to a lack of pollen vitality. However, pollen vitality levels are
not the sole determinant for pollen donor success. It is not known at what level pollen vitality
becomes a limiting factor for competitive pollinisation.

Paternity analyses
In this study the same method was applied to two different field layouts. In the case of the Gumeracha
site the analysis worked well with the smaller number of cultivars in the grove. The NOVA site had a
large number of cultivars present so that a wider range of pollen donors could be identified for the
mother trees. However, the increase in potential pollen donors presented a greater challenge for the
marker system to identify the paternal parent. For 64% of the embryos studied, pollen parents having
LOD scores above the threshold were identified after the analyses. For the remaining 36 % of the
embryos analysed the LOD scores were below the threshold range and pollen donors could not be
identified. The application of more markers to these embryos would help to resolve the pollen parents.

Likely cross-compatibilities found at Gumeracha

Top likely pollen donors identified at the Gumeracha site for the following mother trees were:
• Frantoio: Kalamata, Mission
• Kalamata: Frantoio, Koroneiki.
• Koroneiki: Mission, UC13A6, Kalamata.
• Mission: Koroneiki.
• Barnea: Kalamata, Mission, Pendolino.

An interesting observation at this site was the difference in receptivity to pollen donors for each of the
maternal cultivars within the orchard. The dominant pollinisers were not necessarily the closest
neighbours, indicating that compatibility, rather than proximity, has a major role in pollen success.
This has also been reported in other studies using isozyme markers to identify pollen donors in
orchards of cherry and avocado (Brant et al. 1999; Sulaiman et al. 2004). Barnea was receptive to
several pollen donor cultivars, whereas the majority of Mission embryos were pollinised by
Koroneiki. Twelve of the seventeen genotypes identified in the orchard did not contribute pollen to
the Mission embryos, indicating that Mission may be cross-incompatible with most of the cultivars in
the grove.

The two major pollen donors found in the orchard were Koroneiki and Mission. These two cultivars
had the largest number of trees in the grove, and would therefore have had a favourable bias on the
overall pollen load. Koroneiki appeared to be a strong pollen donor for both Mission and Kalamata.
However, the fact that none of the embryos from either Frantoio or Barnea were assigned to
Koronieki, even though the mother trees were near the Koroneiki block, indicates possible
incompatibility between these cultivars. In contrast, Mission, pollinated at least some embryos from
all the cultivars tested. The results confirm the findings of previous reports, that self-incompatibility
and cross-incompatibility commonly occur in olive (Griggs et al. 1975; Moutier et al. 2001; Wu et al.
2002; Lavee et al. 2002).

Pollen donors for a significant proportion of the embryos tested could not be identified. It is proposed
that the mother trees were pollinated either by pollen from unidentified cultivars in the orchard or by

36
wind-borne pollen from trees outside the orchard. Analysis of the paternal alleles identified in these
embryos revealed a similar contribution in both the years and indicated that more than one unknown
genotype was involved.

Likely cross-compatibilities found at the NOVA site

Top likely pollen donors identified at the NOVA site for the following mother trees were:
• Frantoio: Kalamata, Coratina, Hardys Mammoth, Manaiki, Picual/Azapa
• Kalamata: Koroneiki, Frantoio/WA Mission, Barnea
• Koroneiki: Mission, Hojiblanca, Hardy’s Mammoth, Arecuzo
• Mission: Koroneiki, Arbequina, Amelon, Blanquette, Oblonga

There were notable similarities between the results obtained at both sites. At least one of the major
likely pollen donors identified for the mother trees at Gumeracha was also identified as a major likely
donor for the trees at the NOVA site. This result indicates that these cultivars may be effective pollen
donors in more that one environment. The fact that they are likely pollen donors at both sites also
indicates that they are potentially efficient donors for the respective mother trees. In the NOVA site
cultivars such as Koroneiki, Kalamata and Mission are not as highly represented in the grove as they
are at Gumeracha, which shows that they are likely to be the preferred pollen donors for the mother
trees.

Similarly some potential cross-incompatibilities were evident at both sites. Koroneiki only pollinised
2 embryos from Frantoio, whereas it was a significant donor to other mother trees. Conversely,
Frantoio only pollinised one embryo from Koroneiki at NOVA and none at Gumeracha indicating
reciprocal cross-incompatibility between these two cultivars. Kalamata and Mission also exhibited
reciprocal cross-compatibility at both sites.

Selfing
Another point of interest arising from the results is that the olive trees tested at Gumeracha rarely self-
fertilised. Only two cases of self-pollination were observed out of the 800 embryos tested, and both of
these were from Mission. It has previously been reported that self-pollination in Mission resulted in
very low fruit set (Griggs et al. 1975). The low incidence of self-fertilization was not unexpected as
olives are predominantly allogamous, exhibiting high levels of heterozygosity and DNA
polymorphism among individuals (Rallo et al. 2000).

In contrast we found an increase in self-fertilised embryos in both Koroneiki and Frantoio at the
NOVA site. Mission and Kalamata had a low incidence of selfing with 3% and 2% of embryos
identified as selfed, respectively. Frantoio had 7% and Koroneiki, 15% of embryos as being products
of selfing. The environment may have affected the self-incompatibility response. Koroneiki has been
reported previously to be self-compatible (Lavee et al. 2002), and Androulakis and Loupassaki (1990)
found Koroneiki to be partially self-fertile. It has been shown that self-incompatibility breaks down
due to an increase in temperatures in sweet cherry (Choi and Andersen 2005), lily (Suzuki et al. 2001)
and perennial ryegrass (Wilkins and Thorogood 1992) and this may also be the case in olives.

37
Conclusions
The study identified compatible pollen donors for five commercially important olive cultivars
growing under natural conditions, thus avoiding the adverse effects of artificial cross-pollination.
Bloom time and pollen viability varied between cultivars and years and may have affected the ability
of a cultivar to contribute pollen to the pollen load. The results obtained are a valuable source of
information for planning new olive orchards for optimum yield. Continued research in this area is
needed to offer growers more guidance with orchard design. The identification of the incompatibility
mechanism acting in olives would facilitate compatibility studies through the use of molecular
markers for incompatibility genes.

Implications
The olive industry will benefit from the results of this project through increased productivity. Many
orchards have been established without due regard to the requirements of cross-pollination. These
results will be valuable guidelines to anyone planning new plantings. In addition, they can be used to
increase productivity in established groves through the addition of highly compatible pollen donors.
The effect of introducing compatible pollen donors to a mono-varietal grove could improve yield
(fruit set) from less than 1% up to 4%. New trees can be introduced through grafting onto old trees or
inter-planting. This research also has environmental benefits as a significant increase in yields could
be achieved without expansion of land use, irrigation or addition of fertilisers or chemicals for pest
and disease control.

The impact of different pollen donors on oil quality was not specifically addressed in this research
project. This impact of pollen donors on characteristics of the fruit is called metaxenia. Metaxenia is
only found in a few species and is not common to all fruit and nut crops. Griggs et al (1975) showed
that there is no physical effect of the pollen donor on fruit size or shape which indicates that
metaxenia does not occur in olives. A comprehensive study would require paternity testing of fruit
samples combined with individual fatty acid testing of the oils in the fruit flesh to determine if there is
any effect.

38
Recommendations
Planting recommendations
To compensate for the variability in pollen viability and flowering times between seasons it is
suggested that more than one pollen donor cultivar is present in a commercial grove, preferably three
or four.

The following combinations of mother trees and pollen donors were found to be compatible under the
conditions of this study:

Mother tree Recommended Pollen Donor

Frantoio Kalamata, Mission, Coratina
Kalamata Frantoio, Koroneiki, Barnea
Koroneiki Mission, Hojiblanca
Mission Koroneiki, Arbequina
Barnea Kalamata, Mission

Plantings found to be incompatible under the conditions of this

study
Single cultivar blocks are not recommended due to the low incidence of selfing in most olive
cultivars.

The following combinations of mother trees and pollen donors were found to be incompatible under
the conditions of this study.
Frantoio and Barnea
Kalamata and Mission
Mission and Kalamata
Kalamata and Manzanillo
Koroneiki and Frantoio

Verdale (SA) is not recommended as a pollen donor due to frequent male sterility, observed in this
study.

Future research
It is strongly recommended that continued research be conducted in the area of pollen compatibility in
olives. Further work is needed to understand the genetic system operating in olives, including the
identification of the genes involved. Information on compatibilities between more cultivars will assist
in orchard planting designs to maximize fruit yields.

39
References
Androulakis II, Loupassaki MH (1990) Studies on self-fertility of some olive cultivars in the area of
Crete. Acta Horticulturae 286: 159-162.
Antognozzi E, Standardi A (1978) Studio della biologia fiorale negli olivi 'Gentile di Chieti' e 'Dritta
di Moscufo'. Rivista-della-Ortoflorofrutticoltura-Italiana 62: 461-469.
Ateyyeh AF, Stosser R, Qrunfleh M (2000) Reproductive biology of the olive (Olea europaea L.)
Cultivar 'Nabali Baladi'. Journal of Applied Botany - Angewandte Botanik 74: 255-270.
Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, Albright LM,
Coen DM, Varki A (1987) Current protocols in molecular biology. John Wiley & Sons, Inc.
U.S.A.
Bartolini G, Prevost G, Messeri C (1994) Olive tree germplasm: Descriptor lists of cultivated varieties
in the world. Acta Horticulturae 356: 116-118.
Bartoloni S, Guerriero R (1995) Self-compatibility in several clones of oil olive cv. Leccino.
Advances Horticultural Science 9: 71-74.
Bini G, Lensi M (1981) Observazioni su alcuni aspetti dell 'ontogenesi fiorale nell 'olivo. Rivista-
della-Ortoflorofrutticoltura-Italiana 65: 371-380.
Booth N, Davies G (1996) Olive growing in southern Australia. Primary Industries South Australia
(PISA).
Bradley MV, Griggs WH, Hartmenn HT (1961) Studies on self- and cross- pollination of olives under
varying temperature conditions. California Agriculture: 4-5.
Brant B, Witherspoon J, Granger AR, Collins GG (1999) Identification of pollen donors for the sweet
cherry cultivars 'Stella' and 'Summit' by isozyme analysis. Australian Journal of Experimental
Agriculture 39: 473-477.
Buteler MI, LaBonte DR, Jarret RL, Macchiavelli RE (2002) Microsatellite-based paternity analysis
in polyploid sweetpotato. Journal of the American Society for Horticultural Science 127: 392-
396.
Chaix G, Gerber S, Razafimaharo V, Vigneron P, Verhaegen D, Hamon S (2003) Gene flow
estimation with microsatellites in a Malagasy seed orchard of Eucalyptus grandis. Theoretical
and Applied Genetics 107: 705-712.
Choi C, Andersen RL (2005) A preliminary study on physiological and S-allele-specific breakdown
of self-incompatibility in sweet cherry. Acta Horticulturae 667: 83-87.
Cipriani G, Marrazzo MT, Marconi R, Cimato A, Testolin R (2002) Microsatellite markers isolated in
olive (Olea europaea L.) are suitable for individual fingerprinting and reveal polymorphism
within ancient cultivars. Theoretical and Applied Genetics 104: 223-228.
Connell JH (1994) History and scope of the olive industry. University of California, Division of
Agriculture and Natural Resources, Oakland, CA, California, 1-9 p.
Cuevas J, Diaz-Hermoso AJ, Galian D, Hueso JJ, Pinillos V, M. P, Sola D, Polito VS (2001)
Response to cross pollination and choice of pollinisers for the olive cultivars (Olea europaea
L.) 'Manzanilla de Sevilla', 'Hojiblanca' and 'Picual'. Olivae 85: 26-32.
Cuevas J, Polito VS (1997) Compatibility relationships in 'Manzanillo' olive. HortScience 32: 1056-
1058.
Cuevas J, Rallo L (1990) Response to cross-pollination in olive trees with different levels of
flowering. Acta Horticulturae 286: 179-182.
Cuevas J, Rallo L, Rapoport HF (1994) Initial fruit set at high temperature in olive, Olea europaea L.
Journal of Horticultural Science 69: 655-672.
Dal Pero Bertini GV (1960) Olive growing and processing. Commonwealth Scientific and Industrial
Research Organisation, Melbourne.
de la Rosa R, James CM, Tobutt KR (2002) Isolation and characterisation of polymorphic
microsatellites in olive (Olea europaea L.) and their transferability to other genera in the
Oleaceae. Molecular Ecology 2: 265-267.

40
de la Rosa R, James CM, Tobutt KR (2004) Using microsatellites for paternity testing in olive
progenies. Hortscience 39: 351-354.
Diaz A, Martin A, Rallo P, Barranco D, De la Rosa R (2006) Self-incompatibility of 'Arbequina' and
'Picual' olive assessed by SSR markers. Journal of the American Society for Horticultural
Science 131: 250-255.
El-Kholy M (2001) Olive fruit set - How it really happens. The Olive Press: 14-16.
Ellstrand NC, Devlin B, Marshall DL (1989) Gene flow by pollen into small populations: Data from
experimental and natural stands of wild radish. Proceedings of National Academy of Science,
USA 86: 9044-9047.
Fernandez-Escobar R, Gomez-Valledor G (1985) Cross-pollination in 'Gordal Sevillana' olives.
HortScience 20: 191-192.
Fernandez-Escobar R, Gomez-Valledor G, Rallo L (1983) Influence of pistil extract and temperature
on in vitro pollen germination and pollen tube growth of olive cultivars. Journal of
Horticultural Science 58: 219-227.
Fontanazza G, Rugini E, Mencuccini M (1980) Ricerca di idonei impollinatori delle cv. Ascollana
Tenera e Giarraffa. Annali della Facolta di Agraria, Perugia 34: 119-133.
Gerber S, Mariette S, Streiff R, Bodenes C, Kremer A ( 2000a) Comparison of microsatellites and
amplified fragment length polymorphism markers for parentage analysis. Molecular Ecology 9:
1037-1048.
Ghrisi N, Boulouha B, Benichou M, Hilali S (1999) Agro-physiological evaluation of the
phenomenon of pollen compatibility in olive. Case of the Mediterranean Collection at the
Menera Station, Marrakech. Olivae 79: 51-59.
Griggs WH, Hartmann HT, Bradley MV, Iwakiri BT, Whisler JE (1975) Olive pollination in
California. Division of Agricultural Sciences, University of California, California, U.S.A., 50 p.
Griggs WH, Hartmann HT, Bradley MV, Iwakiri BT, Whisler JE (1975) Olive pollination in
California. California Agricultural Experimental Station, Bulletin 869.
Guerin JR, Sweeney SM, Collins GG, Sedgley M (2002) The development of a genetic database to
identify olive cultivars. Journal of the American Society for Horticultural Science 127: 977-
983.
Hartmann HT, Opitz KW (1966) Olive production in Callifornia. Division of Agricultural Sciences,
University of California, California, U.S.A., 64 p.
Isagi Y, Kanazashi T, Suzuki W, Tanaka H, Abe T (2004) Highly variable pollination patterns in
Magnolia obovata revealed by microsatellite paternity analysis. International Journal of Plant
Sciences 165: 1047-1053.
Jackson JF, Clarke GR (1991) Gene flow in an almond orchard. Theoretical and Applied Genetics 82:
169-173.
Khadari B, Breton C, Moutier N, Roger JP, Besnard G, Bervillé A, Dosba F (2003) The use of
molecular markers for germplasm management in a French olive collection. Theoretical and
Applied Genetics 106: 521-529.
Lavee S (1998) Pollination and cross pollination requirements of olive cultivars. The Tree Cropper
13: 14-15.
Lavee S, Datt Z (1978) The necessity of cross pollination for fruit set of Manzanillo olives. Journal of
Horticultural Science 53: 261-266.
Lavee S, Rallo L, Rapoport HF, Troncoso A (1996) The floral biology of the olive: Effect of flower
number, type and distribution on fruitset. Scientia Horticulturae 66: 149-158.
Lavee S, Taryan J, Levin J, Haskal A (2002) The significance of cross-pollination for various olive
cultivars under irrigated intensive growing conditions. Olivae 91: 25-36.
Lewis D (1994) Self-incompatibility in woody horticultural species. In: E. G. Williams AEC, R. B.
Knox. (ed) Genetic Control of Sef-incompatibility and Reproductive Development in
Flowering Plants. Kluwer Academic publishers, Dordrecht; Boston; London, pp 88-101.
Lian C, Miwa M, Hogetsu T (2001) Outcrossing and paternity analysis of Pinus densiflora (Japanese
red pine) by microsatellite polymorphisnm. Heredity 87: 88-98.

41
Marco Ld, Giovannini D, Marra FP, Viglianisi G (1990) Reproductive and vegetative behaviour of
four table olive cultivars. Acta Horticulturae 286: 187-190.
Martin GC (1990) Olive flower and fruit population dynamics. Acta Horticulturae 286: 141-153.
Martin GC (1994) Botany of the olive. In: Ferguson L, Sibbett SG, Martin GC eds) Olive Production
Manual. University of California, Division of Agriculture and Natural Resources, Oakland, CA,
California, pp 19-21.
Mekuria GT, Collins GG, Sedgley M (1999) Genetic variability between different accessions of some
common commercial olive cultivars. Journal of Horticultural Science and Biotechnology 74:
309-314.
Mookerjee S, Guerin J, Collins G, Ford C, Sedgley M (2005) Paternity analysis using microsatellite
markers to identify pollen donors in an olive grove. Theoretical and Applied Genetics 111:
1174-1182.
Moutier N (2000) Self-fertility and inter-compatibilities of sixteen olive varieties. In: Vitagliano C,
Martelli GP (eds) Fourth International Symposium on Olive Growing. International Society of
Horticultural Science, Bari, Italy, pp 209-211.
Moutier N, Garcia G, Féral S, J.-C. S (2001) Pollination control in olive orchards. Olivæ 86: 35-37.
Oreilly P, Herbinger C, Wright J (1998) Analysis of parentage determination in Atlantic salmon
(Salmo salar) using microsatellites. Animal Genetics 29: 363-370.
Pinney K, Polito VS (1990) Olive pollen storage and in vitro germination. Acta Horticulturae 286:
207-210.
Polito VS, Aradhya M, Dangl J, Grant J, Pinney K, Simon C, Vaknin Y, Weinbaum S (2003) Walnut
pollination dynamics: Pollen flow and pollen loads in walnut orchards. Hortscience 38: 741.
Radi AA, Caballero JM, Rallo L (1990) Influence of fruit set, rate of pistillate inflorescences and
level of flowering in the fruiting of olive bearing shoots. Acta Horticulturae 286: 155-158.
Rallo L, Cuevas J, Rapoport HF (1990) Fruit set pattern in self and open pollinated olive cultivars.
Acta Horticulturae 286: 219-222.
Rallo L, Martin GC, Lavee S (1981) Relationship between abnormal embryo sac development and
fruitfulness in olive. Journal of the American Society for Horticultural Science 106: 813-817.
Rallo P, Dorado G, Martin A (2000) Application of microsatellite markers in olive breeding. Acta
Horticulturae 586: 60-62.
Rapoport HF, Rallo L (1990) Ovule development in normal and parthenocarpic olive fruits. Acta
Horticulturae 286: 223-226.
Reale L, Sgromo C, Bonofiglio T, Orlandi F, Fornaciari M, Ferranti F, Romano B (2006)
Reproductive biology of olive (Olea europaea L.) DOP umbria cultivars. Sexual Plant
Reproduction 19: 151-161.
Robledo-Arnuncio JJ, Gil L (2005) Patterns of pollen dispersal in a small population of Pinus
sylvestris L. revealed by total-exclusion paternity analysis. Heredity 94: 13-22.
Rohlf FJ (1993) Numerical taxonomy and multivariate analysis system. In: Department of Ecology
and Evolution, State University of New York, Exeter Software, Setauket, New York, New
York.
Rovira M, Tous J (2000) Pollen viability in several Arbequina olive clones. In: Vitagliano C, Martelli
GP (eds) Fourth International Symposium on Olive Growing. Bari, Italy, pp 197-200.
Sanz-Cortes F, Martinez-Calvo J, Badenes ML, Bleiholder H, Hack H, Llacer G, Meier U (2002)
Phenological growth stages of olive trees (Olea europaea). Annals of Applied Biology 140:
151-157.
Sedgley M (1994) Self-incompatibility in woody horticultural species. In: E. G. Williams AEC, R. B.
Knox. (ed) Genetic Control of Sef-incompatibility and Reproductive Development in
Flowering Plants. Kluwer Academic publishers, Dordrecht; Boston; London, pp 141-163.
Sefc KM, Lopes MS, Mendonca D, Rodriguez dos Santos M, Laimer da Camara Machado M, da
Camara Machado A (2000) Identification of microsatellite loci in olive (Olea europaea) and
their characterisation in Italian and Iberian olive trees. . Molecular Ecology 9: 1171-1193.

42
Sharma PC, Thakur DR, Sharma MR (1976) Pollination and fruit development studies in olive, Olea
europaea. Food Farming and Agriculture 8: 24-26.
Sibbett SG, Freeman M, Ferguson L, Polito VS (1992) Effect of topically applied 'Sevillano' pollen
on normal seeded and parthenocarpic "shotberry" fruit set of 'Manzanillo' olive.
HortTechnology 2: 228-230.
Singh RP, Kar PL (1979) Compatibility Studies in Some Olive Cultivars. Himachal Pradesh Journal
of Agriculture: 8-15.
Singh RP, Kar PL (1980) Compatibility studies in some olive cultivars. Progressive Horticulture 12:
9-15.
Steenkamp J, Wiid I, Lourens Pv, Helden PV (1994) Improved method for DNA extraction from Vitis
vinifera. American Journal of Enology and Viticulture 45: 102-106.
Streiff R, Ducousso A, Lexer C, Steinkellner H, Gloessl J, Kremer A (1999) Pollen dispersal inferred
from paternity analysis in a mixed oak strand of Quercus robur L. and Q. petraea (Matt.) Leibl.
Molecular Ecology 8: 831-841.
Sulaiman Z, Collins G, Witherspoon J, Sedgley M (2004) Identification of pollen donors for the
avocado cultivar Gwen in a mixed orchard by isozyme analysis. . Journal of Horticultural
Science and Biotechnology 79: 571-575.
Suzuki H, Tsuruhara A, Tezuka T (2001) Regulations of the C2H4-forming system and the H202-
scavenging system by heat treatment associated with self-incompatibility in lily. Sexual Plant
Reproduction 13: 201-208.
Thompson EA, Meagher TR (1987) Parental and sib likelihood in geneology reconstruction.
Biometrics 43: 585-600.
Tsalikidis I, Voyiatzi C, Tamoutseli K (1999) The use of the olive tree in ancient and contemporary
garden design of the Mediterranean region. Acta Horticulturae 474: 767-770.
Ugrinovic K, Stampar F (1996) Fertilization of olive (Olea europaea L.) cultivars 'Istrska Belica',
'Pendolino' and 'Leccino' by different pollinators. Zbornik Biotehniske Fakultete Univerze v
Ljubljani, -Kmetijstvo 67: 183-188.
Villemur P, Musho US, Delmas JM, Maamar M, Ouksili A (1984) Contribution a l 'etude de la
biologie florale de l'oliver (Olea europaea L.): sterilite male, flux pollinique et periode effective
de pollinisation. Fruits 39: 467-473.
Voyiatzi C, Tamoutseli K, Voyiatzis D, Bakirtzi O (1999) The olive tree in ancient Greek life and art.
Acta Horticulturae 474: 763-766.
Weisman Z, Avidan N, Lavee S, Quebedeaux B (1998) Molecular characterisation of common olive
varieties in Israel and West Bank using Random Amplified Polymorphic DNA (RAPD)
markers. Journal of the American Society for Horticultural Science 123: 837-841.
Wilkins PW, Thorogood D (1992) Breakdown of self-incompatibility in perennial ryegrass at high-
temperature and its uses in breeding. Euphytica 64: 65-69.
Wu S (2002) Sexual compatibility and construction of molecular linkage maps in olives (Olea
europaea). In: Department of Horticulture, Viticulture and Oenology. University of Adelaide,
Adelaide.
Wu SB, Collins G, Sedgley M (2002b) Sexual compatibility within and between olive cultivars.
Journal of Horticultural Science & Biotechnology 77: 665-673.

43