Home

Pollination and seed set

Tree Autecology and Biology

Temperate Forest Ecology

Tropical Forest Ecology

Forestry

Sand Dune Ecology

East Usambaras

Pitcairn Islands

Most maples are insect pollinated though a few quantitative studies have shown that seed set occurs without the help of insect vectors. For instance in Acer saccharum Gabriel & Garrett (1984) have shown that wind alone is sufficient to produce a normal seed set, whereas in A. pensylvanicum no wind pollination occurs (Sullivan 1983), and apomixis is an important means of reproduction in A. spicatum (Sullivan 1983).

Sycamore is commonly described as insect pollinated and Mathews (1964) states that good flying weather, that is bright, dry and fairly calm weather for insects is required for pollination. Brian (1957) found that Bombus lucorum is a frequent visitor of sycamore and B. pratorum has also been observed on flowers. B. hortorum will not visit the flowers if the former two species are present but has been observed in large numbers on a flowering sycamore on the Isle of Coll (Heslop-Harrison 1939). Svobodová  (1974) noted that Bombus species visited male flowers more often than female flowers. Acer pollen is important to honeybees for at least one week in Spring (Synge 1947). Chambers (1946) observed that Andrenas armata after collecting pollen from fruit trees, changed constancy to sycamore, when it came into flower. Proctor & Yeo (1973) recorded visits to flowers by sawflies and Bibio hortulanus (Diptera). A list of insects visiting sycamore flowers may be found in Knuth (1908, p 257). Most of these insects may play a role in the pollination of sycamore. Another species of insect Dilophus febrilis is widely distributed on the north coast of Ireland and often abundant in May (D'Arcy-Burt & Chandler 1987) and I have often seen it on sycamore flowers in large numbers. D'Arcy-Burt & Blackshaw (1991) stated that Dilophus febrilis is sometimes attracted in vast numbers to feed on honeydew produced by aphids but it is probable that on sycamore flowers I observed they would feed on nectar. It is known to be a good pollinator of apple trees and might be an important pollen vector of sycamore.

Isolated trees carry regular fruit crops. It is therefore possible that 

1. selfing occurs, 
2. pollinators travel long distances, 
3. apomixis occurs or 
4. flowers are wind-pollinated. 

Although Hagman (1975) stated that the Aceraceae appear to have gametophytic control of incompatibility, Piatnitsky (1934) showed that self-pollination is high (54% of the 125 investigated flowers produced fruits probably from the same tree) in controlled pollination experiments (flowers isolated in an `envelope of parchment paper' for a month which was removed after two weeks and an open anther from the same tree was applied on the receptive stigma). Although Svobodová  (1974) thought that self-pollination is practically impossible because of differences in anthesis between male and female flowers, in isolated individuals self-pollination is possible as long as the flowering of all inflorescences is not synchronized. Published evidence suggests that wind pollination is important. Piatnitsky (1934) and Semm (1965) have shown that wind pollination is possible and indirect evidence such as the high proportion of Acer atmospheric pollen in Britain (Hyde 1950, Hyde & Williams 1961) and in Irish pollen diagrams support this view (K.J. Edwards 1985, Hall 1990). Furthermore, Hesse (1979) classified the species as amphophilous according to its "Pollenkitt". This section investigates the relative importance of wind and insect pollination in twenty trees at Murlough Bay, Co. Antrim.

METHODS

The Murlough Bay Breesha plantation was selected for the study of wind-pollination for the following reasons: 

1. its remoteness and hence decreased likelihood of vandal damage, 
2. a large number of trees with an unbiased "sex ratio", 
3. the unexposed position of most of the site limiting strong wind damage problems, and 
4. yearly and usually prolific flowering of most trees.

Out of a population of 37 trees, it was possible to select ten protogynous and ten protandrous trees bearing a large enough number of inflorescences close to the ground. The selection also included trees of contrasting flushing dates and varying exposure to wind.

On each tree three flowering branchlets were randomly chosen around the base of the canopy (2-4m above ground). The process involved either jumping, or climbing a step ladder and the first branchlet reached by the left hand was selected. On each branchlet 10 inflorescences were tagged. Every other inflorescence had a pollination bag attached around it in order to prevent insect pollination. The trees were visited twice a week in May and June 1991 and the bags were fitted prior to anthesis. The bags were 15cm long and 8cm across and were made of nylon curtain material with a mesh of about 1mm. The number of flowers, their position, the number of carpels and the time of anthesis was recorded. 

The possible occurrence of agamospermy was checked on one protandrous and one protogynous tree using large brown paper bags fitted with a small plastic window. These bags were fitted around small branchlets inclusive of the leaves of two trees and the inflorescences were emasculated. The same observations as above were made, but no seed set was observed. Casual observations on insect visiting flowers were also made.

On the 24th of September 1991, the branches were collected and brought back to the laboratory, where the number of fruits on each infructescence and the number of empty samaras were recorded.

Of the nine infructescences lost during the experiment, two were a result of goat and/or cattle grazing, four broke off during transport and three were not relocated in the autumn.

The wind conditions at the study site were investigated on windy occasions in November and December 1991 at eight locations representing contrasting degrees of exposure. Relative wind speed was estimated from anemometers attached to poles at a height of 1.5m above the ground. Weather data for May and June 1991 from Coleraine were used for the determination of the wind strength and direction at the time of female anthesis.

RESULTS

As shown in Table 1 significant differences (p<0.05) in seed set were found between sexual morph (F-ratio = 76.9, df 1,116), and between wind pollination and control (F-ratio = 130.7, df 1,116). Protogynous individuals average seed set in control (insect + wind pollinated 58.7%) and wind pollinated (11.7%) flowers were much higher than in protandrous trees (seed set 19.1% and 2.9% respectively). In protogynous individuals seed set in control conditions varied from 40.5 to 80.6% and in protandrous trees from 0% to 47.7%, whereas under wind pollinated conditions protogynous tree seed set varied from 1.7% to 30.2% and that of protogynous trees between 0% and 15.3%. Among protogynous trees differences in seed set between the two pollination treatments were significant in eight out of ten trees whereas in the other sexual morph it was only in one tree.

Table 1. Relative importance of insect + wind (control) and wind pollination in seed set in protogynous and protandrous sycamore at Murlough Bay, Co. Antrim. (Sign. level = significance level, NS = not significant, * = p<0.05, ** = p<0.01, av. = average)

Protogynous

Protandrous

Wind Speed

The relationship between wind pollination and natural conditions (wind + insect pollination) is illustrated in Fig. 1. It shows that wind pollination has an important impact on seed set only in trees with higher seed set (above about 35%). The differences observed among trees with high seed set can readily be explained by their exposure to wind and the wind conditions at the time of female anthesis. The relative wind speeds are given in Table 1. The degree of exposure of trees varied according to wind direction. In the bay, wind direction differed from the general wind pattern particularly because of the effect of the cliffs of Fair Head to the north (see Binggeli & Rushton 1985). All trees with a high wind pollinated seed set were individuals which reached female anthesis during a very windy period. Trees M1, M3, M37 and M2 reached female anthesis when strong northerly winds prevailed in the Bay and all had a high degree of exposure to them. Tree M13, sheltered from northerly winds, reached female anthesis during a period of strong westerly winds. The relative wind speeds under three contrasting wind directions for each tree are given in Table 1. These estimates were obtained when trees had shed their leaves and it is likely that the difference would be greater when trees are in full leaf.

Figure 1. Relation between seed set under wind pollination (wind) and insect + wind pollination (control),  = protandrous and = protogynous individuals.

Whenever flower anthesis was at its peak, particularly that of the male flowers, a very large number of insects were observed on the flowers, particularly Bombus species as previously recorded among others by Heslop-Harrison (1939). An observation on Bombus lucorum, not recorded in the literature, was made on the morning of the 13/6/1991 on the exposed and partly wind-pruned tree M2. The weather was very windy (13 knots at Coleraine and of a northerly direction in the Bay) but sunny with a high wind chill factor. Three Bombus lucorum were found hanging on three separate inflorescences. This protandrous tree was at the peak of its maleI flowering phase, with many stamens having already released their pollen. The bees were found on male flowers, facing upward, wrapped around the inflorescence unable to fly. One of the twigs, including the inflorescence and the bee, was broken off and brought to the ground, however the bee did not react to this treatment. The thorax was expanding and contracting very regularly. 

According to the Coleraine weather data, similar northerly windy conditions (wind speed above ten knots) were observed on seven days during the period of female anthesis (about a month ) indicating that during about a quarter of that period weather conditions were not suitable for bumble bees foraging. 

DISCUSSION

The above results clearly indicate that in trees reaching female anthesis during periods of high winds, wind pollination has a significant impact on seed set. However, wind pollination only occurs in trees which are found in exposed locations. Sycamore pollen does not appear to travel far unless exposed to high winds. In an Austrian forest along a 2km transect Kral & Mayer (1968) found that sycamore pollen represented less than 1% of the total pollen in the top soil, except in the close vicinity of a few adult trees, indicating that pollen mostly falls below flowering trees. Isolated trees growing in exposed conditions can be pollinated by wind dispersed pollen, either from distant trees, or from its own pollen if the flowering sequence of some of its inflorescences overlap.

In a study of the effect of Formica rufa on levels of herbivory in sycamore and related impact on tree growth Whittaker & Warrington (1985) found that "fewer viable seeds were produced from trees with normal herbivore levels than from those with low herbivore damage" as well as a reduction in the mean annual radial growth. Their results are summarised in Table 2. The difference in the percentage of fruits with seeds between two foraged trees carrying the lower herbivore densities and three unforaged individuals was significantly different in both years but no significant differences in average number of fruits were found. Their results also indicate that more samaras contain embryos when a larger fruit crop is produced.

Table 2. Mean fruit fall and percentage of fruits with embryos in sycamore in Lancashire in 1982-1983 (adapted from Table 6 in Whittaker & Warrington 1985).

In both this study and Whittaker & Warrington (1985) large differences in the percentage of filled samaras are observed but the results are explained either by the variation in sex expression or differences in herbivore damage. Dixon (pers. comm.) made a similar suggestion, i.e. large aphid numbers reduced seed production in sycamore. It can be suggested that such differences in interpretation arise from the fact that Whittaker & Warrington (1985), like many other ecologists (e.g. Dixon pers. comm., Pigott & Warr 1989) who have investigated sycamore, do not know or do not take into account the impact of sex expression on various tree characters. Furthermore, they used a rather small sample size of five trees and their estimates of the percentage of filled fruits is obtained from seeds collected on the ground instead from the trees themselves. The data presented in Table 2 suggest that ungrazed tree 1 and 2 are protandrous individuals, Grazed tree 1 and 2 are protogynous and the status of Grazed tree 3 is unclear. However, only observation of flowering or fruiting material can ascertain this. In view of my results and of Whittaker & Warrington's methodological inadequacies, their results are likely to reflect differences between sexual morph and a small sample size. Another possibility is that trees of different sexual morph are selectively foraged by ants resulting from differential distribution of their prey. Support for such an hypothesis is suggested by recent findings pointing out that in Arctic Salix spp. and Populus tremula male individuals tend to be more palatable to herbivores than female individuals (Elmqvist et al. 1988, Hjältén 1992).