The A's & B's of Avocado Flowering

The avocado (Persea americana) is a member of the Lauraceae family which ranks amongst the oldest flowering plants in the fossil record. Archaeological sites in Mexico document avocado consumption as far back as 7-8000 BC, with domestic selection of favourable plants beginning in 4000-2800 BC. From these ancient times our efforts to maximise the quantity and quality of edible produce have been ever-increasing. However for much of recorded history, practices have been based simply on harvesting wild plants or selection/preservation of randomly produced superior trees, through to vegetative propagation to assist in continuation of good product. Real advances over most of this lengthy period were very slow because of a lack of understanding of the underlying biology, and it was also compromised with many practices becoming enshrined in unfounded traditional or culturally-based management techniques. More reliable evidence concerning species behaviour and improvement had to wait largely for the emergence of controlled scientifically-based studies in the last hundred years or so. Horticulture (the study of various crop producing plants) and pomology (the study of fruit producing species) are both now well-established disciplines that have made and continue to make profound contributions, including with avocados, resulting in many-fold increases in yields, quality, improved disease resistance etc - all desperately needed with a world population that continues to increase. Nevertheless, while many properties of avocado fruit trees have now been well characterised, understanding its complicated flowering behaviour has remained enigmatic with much work still needed to be done.

Cross fertilisation in plants confers hybrid vigour and the genetic means to adapt to changes in the environment over evolutionary time. The way in which avocado achieves this, documented in the 1920s, is by diurnally synchronous protogynous dichogamy of the flowers which are perfect, ie have both male and female organs. Flowers bloom over several weeks depending on climate, and they open and close twice, with dichogamy indicating male and female organs are functional at different times; protogynous indicates the female receptive parts become effective before male, and synchronous indicates that A timing is operative to fertilise B, and B to fertilise A. There are two flower types classed as A and B; in A flowers (eg the Hass cv) the female organ is receptive in the morning followed by the male the next afternoon, and in the B type (eg Fuerte), the female is receptive in the afternoon and the male the following morning. Under ideal conditions the overlap between male and female stages then predisposes to cross-pollination, fertilisation, and fruit set between the A and B types, with fruit then proceeding on to maturity. In the absence of convincing evidence that this inter-mixing of cultivar types was necessary for effective crop yields, it nevertheless seemed a logical strategy and was largely practised throughout most of the growing regions in the world from the 1920s onwards. Since then the Hass cv has come to dominate world trade and fetches by far the greatest price, so intermixing plantations with B type pollinators, even if at low ratios, carries a financial cost for growers and complicates management. How beneficial or even necessary is this intermixing, particularly given avocados are now grown in many sub-tropical and tropical regions worldwide that are quite different from avocado origins in Mexico and Central America? It's a flexible species and in Australia can be grown from 17 to 35° S (40° in NZ), ranging from semi-arid winter rainfall conditions to the humid tropics with summer rain.

The avocado is notorious in having profuse flowering (often in the millions) with less than 0.1% of these resulting in harvestable fruit, even with the best cultivars and optimal environmental and cultural conditions. The important questions are – how much of this profound loss is due to inadequate pollination, pollinizer effectiveness, pollen viability, lack or insufficiently rapid pollen tube growth through the style, ovule fertilisation, seed abortion and dysfunctional regulation of plant hormones, insufficient carbohydrate reserves or sunlight, prevailing temperature and humidity history, moisture stress, day length, type of inflorescences, time and duration of anthesis etc? Another unwanted feature of avocado behaviour, especially for Hass, is the tendency for alternate bearing with on and off years, meaning that studies on the influence of varying factors on yield can not be meaningfully interpreted in a single cropping year.

As non-commercial growers, where can we obtain reliable information on the latest findings and recommendations? There are many sources and the most common are personal contacts, popular media programmes, magazines, internet sites and the like. The quality of these can vary enormously. Many are composed by people with little fundamental understanding of the species, and they very often derive content heavily plagiarised from other comparable sources, all perhaps originally stemming from an unreliable source in the first place. The sheer repetition of a similar story all too easily then leads to the perception that with statements from so many sources they must indeed be true. Others may be hyped up by Nurseries just wanting to sell product, or industry organisations sponsoring studies which are dressed up to give an air of respectability but have to be viewed critically because of possible conflict of interest bias. At a much more reliable and objective level than these we have various public agencies such as Agriculture Departments that may produce fact sheets for the public, in addition to what is usually their principal 'raison d'etre' of supporting commercial growers and their associated industries with commercially important species. Sometimes these reports (eg conference papers, patents, yearbooks, annual reports, grower association newsletters etc) may not have been subjected to sufficient and clearly independent peer review, and the underlying science and/or citations for supporting work may not be stated - reasonably enough if intended only for the public. But whenever the underlying science/citations aren't given, the statements and opinions expressed can not be verified. However, an obvious advantage is that content may be more specifically related to local growing regions rather than what applies in other countries. At the highest and most reliable level of evidence sources we have international peer-reviewed studies conducted by experienced researchers working in prestigious universities and institutes with access to the latest technology, fully documenting their methodology and references sources, and published in the best scientific journals.

So jumping over the most common low level information sources, what has the WA Ag Dept advised regarding synchronous dichogamy in avocado orchards and the need for cross-pollination?

In 2001 (Bulletin 4484, Avocado Culture in Western Australia) it was stated:

The mean daily temperature during flowering affects how strictly this pattern is displayed – the higher the temperature the more pronounced it is. Type B varieties are further influenced by temperature, the female organs displaying a very low receptivity to pollen at average temperatures below 18°C during flowering. As a result, type B varieties are less suited to colder climatic regions than type A varieties. Even though avocados are self-fertile, their unusual flowering habit means that you can increase their fruit causing cross-pollination. However under the climatic conditions of the main commercial regions in WA, satisfactory crops can be set on most varieties without the need for cross pollination – provided sufficient pollination vectors are present and active. It is current commercial practice not to plant pollinators when planting a block of the main commercial variety Hass.

The current material on their web site (updated January 2017 but advising content may be out of date and is under review) considers many of the factors mentioned above that can affect yield, including cross-pollination. They again state it's common practice in the south west of WA to have whole plantations consisting only of monocultures, but advise that much more research needs to be done on the importance of cross pollination influencing final yield, with a prime requirement being that there is sufficient overlap in complementary flowering times.

Now let's consider some of the peer-reviewed international literature in chronological order over the last 30 years to give some idea of developments in understanding. Mexico is by far the largest producer in the world, but most of the research has been conducted in the US (major growing regions of Florida and California) with significant contributions from Spain, Australia, South Africa and Israel.

In a 1986 study (Avocado Flowering, Horticultural Reviews, volume 8) it was concluded that:

The events encompassing the induction, flowering, pollination, fertilisation, and post-fertilisation processes are extremely complex. Much of the published information is anecdotal with conclusions sometimes based on results obtained under poorly controlled conditions. The domestic honeybee could not have been involved, since it was not introduced to the new world until the arrival of European colonists. Furthermore, the habits of honeybees are simply not in concert with those of the avocado. Clearly however the honeybee is an effective pollinator of avocado in many areas of the world, but other insects may be as important in pollen transfer in some areas. It is plausible that if cross-pollination does not occur, the avocado has evolved a fail-safe system of self-pollination dependent on environmental conditions.

A 1990 study in Israel (HortScience, 25, 471-473) using enzyme analysis of pollinizer flower type on selfing and outcrossing of fruit in Fuerte (B type) provided evidence re the above conclusion of 'plausibility' of self-pollination, as Fuerte was capable of producing fruit itself. Different pollinizer cvs had a larger effect on both fruit and seed weight than whether pollininzers were type A or B.

A 1994 study in Florida (J Amer Soc Hort Sci, 119, 1200-1207) found that most pollinations in type A and B flowers occurred not during the first opening (ie when female organs became receptive) but during the second male period, as stigmas remained viable and receptive during this second stage. Under the favourable temperatures in Florida of 27-31°C max and 20°C min with humidity averaging 80%, stigmas had not desiccated by the time of the second opening. In a second aspect of the study they found that excluding larger insects such as bees had no effect on pollination rate during the male phase. After testing a number of cvs with similar results, they concluded that the primary means of pollination of commercial cvs in Florida was during the second, male stage of opening.

A 2000 study (Scientia Horticulturae, 86, 135-149) in California over 4 years of cropping using DNA markers investigated the rate of out-crossing in the yield of Hass with Bacon, Fuerte and Zutano as B type pollinizers; in recent years it's been common practice there to plant Hass in monocultures. Two climatic regions with average monthly precipitation and temperatures during bloom of 20mm, max 20°C and min 7°C min, and 70mm and 25°C max and 11°C min were considered. Fuerte was the most effective pollinizer in both climates, and outcrossing rate generally was dependent on climatic region and separation of maternal and pollinizer trees. There was only a weak correlation between yield and outcrossing, and most of this variation appeared due to other causes such as yearly climatic variations. They suggested management techniques such as careful control of water and fertilizer regimes may be as effective as mixed plantings in increasing yields, and these are much less expensive to implement. Self-fertilisation was responsible for a substantial fraction of fruit set, and low yield in California (as elsewhere round the world, with average avocado yields estimated to be only about one third of photosynthetic capacity) was not due to lack of B type pollen sources.

In South East Spain, weather conditions during bloom often allow brief overlaps between female and male phases in a reduced proportion of flowers, and this may explain why growers continue using solid orchards of 'Hass' with no pollinizers. A 2007 study there (Proc VI World Avocado Congress) considered possible pollination vectors. They compared pollen adhesion and resulting fruit set in control flowers freely exposed, to those produced in flowers bagged with either micro perforated plastic (allowing wind and small insects to enter) or with tissue paper (excluding all kinds of insects). Their results showed that only the activity of honey and bumble bees rather than wind allowed fruit set. The activity of thrips caused certain levels of pollen deposition on the stigmas and an initial swelling of avocado flower ovary, however fruitlets soon abscised. The flowers enclosed in paper bags had no pollen grains on the stigmas nor set any fruit. An introduction of commercial hives was therefore recommended to achieve effective avocado pollination in their region.

Another study from California (J. Amer. Soc. Hort. Sci. (2008) 133, 648–652) considered the impact of outcrossing on yields of Hass, which is characterized by excessive flower and fruit abscission resulting in extremely low fruit set; low outcrossing rates could contribute to low yields. The authors hypothesized that self-fertilized flowers and resulting fruit might abscise at a much higher rate than fruit that were the product of outcrossing, even though significant relationships between outcrossing rates and yields have only been established in a few avocado studies. The objective of their research was to investigate the importance of outcrossing on yield in a commercial Hass orchard containing Bacon, an effective pollinizer of Hass. Microsatellite markers were used to determine the rate of outcrossing of fruit persisting through to harvest. Experiments were conducted during sequential on- and off-crop years. They found during both years that outcrossing rates were not related to yield or alternate bearing. They concluded that outcrossing was not the primary factor affecting flower and fruit persistence, and ultimately yield. Similar conclusions were drawn by another research group in Florida (Scientia Horticulturae (2009) 120, 360–366) where complementary flower types were intermixed and they found that a significant portion of the progeny were the result of crosses between like, and not complementary flowering types.

Another southern Spanish study on Hass in 2011 (Ann Appl Biol 158, 354–361) using microsatellite DNA markers considered the effect of separation distance on outcrossing and yield for two successive years between Hass and Fuerte as pollinizer. Outcrossing rate ranged from 74 to 31% and showed a significant decrease as distance increased, but 13 years of data showed no significant differences in Hass yield as distance to the pollen donor trees increased. They suggested that, under their growing conditions, fruit drop in avocado was determined by fertilisation date (later showing increased abscission) and not by embryo genetic composition

Finally, a recent Californian study (Scientia Horticulturae (2016) 199 32–40) considered a number of factors over a two year period that could affect the pollination and fertilization rates of abscising flowers, the viability of developing ovules, and plant hormone concentrations of abscising versus persisting fruit through the first and second fruit drop in a Hass monoculture. Pollination, pollen germination, pollen tube growth and ovule viability of abscising flowers were determined, and in the first drop, cellular deterioration of developing ovules of abscising and persisting fruit was visualized. Abscisic acid (ABA), indole-3-acetic acid (IAA, an auxin), and isopentenyl-adenine (IPA, a cytokinin) concentrations were determined during the on-crop year only. On average ≥ 70% of abscising Hass flowers were pollinated, but due to the failure of the pollen grains to germinate and produce pollen tubes, fertilization never occurred and the ovule deteriorated. Fruit abscission during 1st and 2nd fruit drop was due primarily to a lack of fertilization. Fruit abscission was associated initially with deterioration of the nucellus in the developing ovule in early drop, and subsequently with deterioration of the seed coat in the later drop; both occurred more frequently in abscising than persisting fruit. Abscising fruit had greater ABA concentrations during 1st and 2nd drop compared to persisting fruit, and IAA and IPA were greater in abscising fruit during 2nd drop, but not early drop or late fruit drop. Only 25% of the total number of fruit that abscised per crop year were fertilized. The authors suggested these results provided strong evidence that the majority of flowers and fruit of selfed 'Hass' avocados abscise due to a lack of pollen germination and subsequent fertilization (not pollination). In addition, they suggested that ABA accumulation is related to seed abortion and reduced fruit growth in abscising fruit.

Where does this snapshot of some of the more recent literature leave us with managing our own avocado plants to get good yields? The following points are a summary of most of the stages and factors involved (mentioned in the introductory comments) without giving excerpts of all the additional supporting literature. Although many are somewhat peripheral to our central question of the importance of cross-pollination, they nevertheless all feed in to our ultimate goal of fruit quality and yield, and without recognising their effects can be significant villains contributing to low yield rather than assigning it all to lack of cross pollination.

Barry Madsen

Spacer. Images/Stars Spacer. LitchiLogo