Introduction
Avocado flowers are bisexual, with functional male and female organs that usually operate at different times. This temporal separation is known as synchronous protogynous dichogamy. Each flower first opens in the female stage, when the stigma is receptive, then closes and reopens in the male stage, when the anthers release pollen and the stigma is generally no longer receptive. Cultivars are conventionally classified as type A or B according to the timing of these stages. Type A flowers usually open as females in the morning and reopen as males the following afternoon, whereas type B flowers generally open as females in the afternoon and reopen as males the following morning.
This pattern can promote pollen exchange between cultivars, but it does not operate as a rigid clock under all conditions. Temperature and relative humidity can modify the floral cycle, and flowers at different sexual stages may overlap on the same tree or among trees of the same cultivar. Such overlap helps explain fruit production in single-cultivar blocks. Nevertheless, avocado produces millions of flowers while fewer than 1% typically develop into harvested fruit. This low fruit-to-flower ratio cannot be attributed to pollination alone: pollen delivery and performance, flower quality, conditions during the progamic phase, and competition for resources all contribute to the final outcome.
Research in southern Spain indicates that pollination can be a major limitation, although it is only one part of the process. In earlier experiments, fruit set in hand-pollinated flowers was about fifteen times higher than in flowers left to open pollination, yet substantial flower and fruitlet abscission still occurred after hand pollination. More recent work found that most ‘Hass’ flowers received no pollen during the female stage, and that the probability of pollen germination and tube growth was higher when stigmas received larger pollen loads. These results support a coordinated approach that improves pollen delivery while also considering flower condition, flowering synchrony, environmental conditions and the tree’s capacity to support developing fruit. The higher crop in a high-bearing year was associated mainly with a greater number of flowers, rather than a lower proportion of abscission, underlining the distinction between flower production and the probability that each flower will become a harvested fruit.
Effect of environment
Environmental conditions affect both the timing of flower opening and the processes connecting pollination with fertilization. The effective pollination period is the interval during which a flower retains the capacity to develop into a fruit. It depends on the duration of stigma receptivity, the rate of pollen-tube growth through the style and the longevity of the ovule. These processes are closely linked to temperature and humidity. Consequently, pollen must be delivered at the appropriate floral stage and under conditions that permit germination and fertilization; the presence of pollen alone does not guarantee success.
Experiments under conditions representative of the ‘Hass’ flowering season in southern Spain showed that pollen adhesion and germination were greatest at approximately 20–25°C and high relative humidity in both female- and male-stage flowers. Above 25°C, pollen germination declined, particularly in male-stage flowers, where rapid stigma degeneration may reduce the time available for pollen to germinate. High relative humidity appeared to reduce stigma desiccation and allowed pollen germination under otherwise less favourable conditions. These findings indicate that reproductive performance depends on the microclimate experienced by individual flowers, not simply on whether pollen is available in the orchard.
Temperature also affects pollen-tube growth. Under the experimental conditions, tubes reached the ovary in about four hours at 30°C and eight hours at 20°C. At 10°C, however, pollen tubes were not observed penetrating the ovule even two days after pollination. Faster tube growth at warmer temperatures should not be interpreted as an unconditional benefit, because elevated temperatures can also shorten stigma receptivity and ovule longevity. The relationship between temperature and fertilization is therefore not linear: successful reproduction requires a suitable balance among pollen deposition, germination, tube growth and the period during which the ovule remains viable. In addition, experimental pollinations indicated that pollen deposited during the male stage did not lead to embryo or endosperm development under the conditions tested, emphasizing the importance of effective pollen transfer during the receptive female stage.
Cool conditions can disrupt the usual flowering cycle, especially in type B cultivars, and alter the overlap between male and female stages. In southern Spain, the floral cycle was generally longer at the beginning of the flowering season, when temperatures were cooler, than later in the season. Some overlap between sexual stages provided opportunities for self-pollination or close-pollination even when the standard A/B pattern would predict separation. During the April ‘Hass’ flowering period monitored in the study region, temperatures ranged from 9 to 27°C and relative humidity from 45 to 94%. This variability shows why a single description of the typical flowering cycle can be misleading. Flowering type alone is not sufficient for selecting pollinizers; local temperature patterns and actual flowering synchrony must also be considered.
Warming winters are also changing flowering phenology in southern Spain. Observations at La Mayora indicate that the beginning of ‘Hass’ flowering has advanced by approximately three weeks over the past two decades, with particularly early flowering in 2024. This shift can move flowering into a cooler period, when pollen germination and tube growth are slow and insect activity may be limited. Climate change can therefore affect pollination not only through warmer conditions during flowering, but also by changing when flowers open relative to pollinator activity and suitable temperatures for fertilization.
Flower quality
A flower’s capacity to set fruit is not determined solely by external appearance or by whether pollen reaches the stigma. Flowers that later abscise can look similar to those retained on the tree while differing substantially in internal nutritional status. In avocado, reproductive and vegetative growth occur simultaneously, so flowers and young fruitlets compete with shoots for resources supplied by current photosynthesis and stored reserves. The capacity of an individual flower to develop into a fruit may therefore be partly established before pollination.
Measurements at anthesis revealed considerable variation in ovary starch content among flowers, with only a small proportion showing high levels. Flowers with greater starch reserves were more likely to set fruit. A comparison between flowers opening in April and a later group opening at the end of May, which did not produce fruit under the conditions studied, also found higher starch content in the earlier, more successful group. In hand-pollination experiments, starch content at anthesis was correlated with subsequent fruit set. Because starch in the cortical tissue of the style was related to starch in the ovary, style measurements provided an indirect way to assess variation in floral reserves. Further analyses of boron and carbohydrates, including mannoheptulose, sucrose, perseitol and fructose, also point to differences in internal flower condition. Higher reserves appear to give a flower an advantage, but they are not sufficient on their own: effective pollination and fertilization are still required, and fertilized flowers may subsequently abscise.
The 2021 pollen-load study further showed that the likelihood of pollen germination and pollen-tube growth increased with the number of grains deposited on the stigma. Loads above 20 grains were associated with a greater proportion of flowers showing pollen tubes, whereas loads of five grains or fewer significantly reduced the probability of fruit set. These are trends rather than an absolute threshold: some flowers with relatively few pollen grains set fruit, while others with larger loads did not. Thus, both pollen quantity and flower quality influence the fate of a flower. From a management perspective, the balance between reproductive demand, vegetative flush and available reserves is important during flowering and early fruit development.
Pollinizers
The traditional strategy for improving cross-pollination has been to plant complementary type A and type B cultivars near one another. However, the yield benefit of interplanting complementary cultivars has varied among production regions and studies. Some investigations reported higher yields when pollen donors were nearby, whereas others found a weak or no relationship between distance to a pollen source and yield. This variation is consistent with the sensitivity of flowering behaviour to local conditions and with the fact that pollination is only one of several factors determining fruit retention.
In southern Spain, the proportion of outcrossed fruit on ‘Hass’ trees decreased with increasing distance from a solid ‘Fuerte’ block, but no significant relationship between outcrossing rate and yield was detected in the two years studied. A twelve-year analysis also found no significant effect of distance between complementary cultivars on yield. More recent work nevertheless indicates that fruitlets derived from outcrossing can have a higher probability of reaching maturity than selfed fruitlets. These findings are not necessarily contradictory: a pollen parent can influence the fate of individual fruitlets without producing a consistent increase in total orchard yield. The timing of fertilization, selective fruit drop and competition among developing fruitlets and vegetative growth can all affect the final crop. The economics of pollinizer planting also matter because fruit from some pollinizer cultivars may have a lower market value than ‘Hass’, and the area allocated to them has an opportunity cost.
Flowering overlap is therefore a key criterion when selecting pollinizers. ‘Hass’ and ‘Fuerte’ overlapped for only part of their flowering periods in southern Spain, leaving some of the ‘Hass’ flowering season without ‘Fuerte’ pollen. A cultivar that appears complementary based on its A/B classification may be of limited value if it flowers too early or late. Abnormal or delayed cycles can extend overlap between sexual stages and create additional opportunities for self-pollination, but these responses are environment
dependent. The most useful pollinizers should be selected using local flowering records, overlap of male and female stages and pollination performance; cultivars with fruit quality and market characteristics similar to ‘Hass’ may also reduce the commercial cost of interplanting.
Pollinators
The large increase in fruit set after hand pollination compared with open pollination shows that pollen delivery can limit production. Recent field work in southern Spain provides direct evidence of inadequate pollination under commercial conditions. Across a seven-season study involving 17,288 flowers, the proportion of female-stage flowers with pollen on the stigma averaged 2.85% without introduced honeybee hives, 4.34% with 10 hives per hectare and 13.79% with 24 hives per hectare. Early fruit retention varied between years: the average at the end of June was 0.13%, 0.11% and 0.21%, respectively. The increase in pollen deposition was clearer than a uniform response in fruit retention, illustrating the influence of seasonal variation and other constraints. In five commercial ‘Hass’ orchards, pollen deposition was also generally low and variable despite the presence of pollen-donor cultivars and managed honeybees. These findings indicate that pollination can be a primary limitation, particularly under Mediterranean conditions, while also showing that hive presence alone does not ensure adequate pollen transfer.
Honeybees remain an important tool, but their effectiveness depends on abundance, activity and the timing of visits relative to the receptive female stage. In the seven-year study, increasing the number of hives to 24 per hectare significantly increased the percentage of flowers receiving pollen compared with lower-density treatments. However, pollen deposition did not differ significantly among rows located at increasing distances from the hives, up to approximately 104 m. This suggests that, under those orchard conditions, hives need not all be concentrated at one point, although optimal density and distribution should be tested locally. Pollen deposition also changed during the flowering season and generally increased towards its end, alongside rising temperatures. Because bee activity can decline during cool weather, an early flowering period may be particularly vulnerable even when hives are present.
A broader pollination strategy should combine sufficient managed honeybees with measures that support pollinator diversity. Wild bees, bumblebees, flies such as syrphids, beetles and butterflies may visit avocado flowers, but their contribution to effective pollination varies and should not be assumed from visitation alone. Maintaining suitable ground cover and nesting resources, and avoiding harmful insecticide applications during flowering, can help sustain beneficial insects; ground-cover management should nevertheless be adapted to local orchard conditions. Pollinator activity must be evaluated alongside flower stage, pollen quantity, temperature, humidity and flower quality. The goal is not simply to maximize insect numbers or cross-pollination, but to increase the chance that sufficient pollen reaches receptive flowers at a time when fertilization can proceed and the tree can support developing fruit. Because most detailed evidence summarized here comes from southern Spain, pollinizer arrangements and pollinator-management practices should be validated under the environmental and production conditions of each region.