Phase change in an invasive population
In woody plants saplings do not flower, and the duration of this juvenile period increases as plant lifespan lengthens (Harper & White 1974). The transition from the juvenile phase to one of flowering (adult or mature phase) is called phase change and has been widely reviewed (e.g. Wareing & Black 1959, Matthews 1963, Jackson & Sweet 1972, Zimmerman 1972, Hackett 1985). The relative importance of controlling factors in the transition from the juvenile to the adult stage in woody plants has been much debated. Some authors (e.g. Hackett 1985, Wareing 1987) support the view that phase change occurs when a certain size has been attained, although it is not clear whether it is size per se or a correlated variable which is the controlling factor. Lacey (1986) concludes that both size and age influence time of first reproduction and that resource allocation may also be important, while Lines (1977) notes that colonizing species, such as birch, which are often short-lived begin to flower earlier than long-lived late successional species. Others writers (Salisbury & Ross 1978, Metzger 1988) suggest that the timing of phase change will vary according to certain environmental conditions such as photoperiod or thermoinductive temperatures.
Phase change induces stable differences in gene expression which are transmitted through repeated life cycles of cell divisions. Cuttings from flowering shoots grafted on to juvenile branches will carry on flowering but reversion to a juvenile state may occur under special conditions (Wareing 1987).
Apart from flowering, clear differences between juvenile and adult phases are frequently distinguishable by differences in various vegetative characters, such as leaf shape and size, phyllotaxy, pigmentation, etc. (Kozlowski 1971, Wareing 1987).
In forestry the age at which flowering starts is important for two reasons. Firstly in tree breeding it is beneficial if trees flower early and techniques have been devised to induce precocious flowering (Zimmerman 1972, Horell et al. 1990, Philipson 1990), whereas in tree species where flowering affects tree architecture late flowering is sought for better timber production. For instance in many tree species, including sycamore, flowering is terminal and induces forking of branches including that of the leading shoot. This of course reduces the length of the harvestable stems and decreases the value of the crop as is in the case of sycamore plantations in Denmark (Kjölby et al. 1958, Schaffalitzky de Muchadell 1959).
In deciduous trees the age at which flowering is first observed is not well known and in Europe most recent authors (e.g. Gordon & Row 1982, Metzger 1988, Philipson 1990, Savill 1991) rely on earlier qualitative assessments (e.g. Matthews 1955, Aldhous 1972, Schopmeyer 1974). For sycamore the age of first flowering is given as 15-20 years (Metzger 1988) and that of minimum seed bearing age (Gordon & Row 1982, Evans 1984) or the first good seed crop as 25 to 30 years. This suggests that flowering intensity increases following first flowering, but quantitative data are lacking and little indication is given as to what happens when trees grow older. Evans (1984) stated that seed production in sycamore begins to decline after the age of 70 years. Here, phase change in sycamore is investigated from an invasive population of a birch-rowan-hazel wood. Impacts of flowering on tree architecture are also reported.
METHODS
Thirty-five trees were selected in an invading population of a semi-natural birch-rowan-hazel woodland to include a range of girths at breast height (GBH). The woodland is situated at Murlough Bay, Co. Antrim (specific site IGR D1842) and its ecology, including the pattern of sycamore invasion, is reported in Binggeli & Rushton (1985). Starting on the uphill side of the crown four branches from each tree were cut at mid-height at 90° interval around the canopy and brought back to the laboratory. Most of the trees were sampled in November 1988 except for the five largest individuals collected in October 1989. Girth was measured at breast height (GBH) except for trees forking below that height, in these instances the girth was recorded under the split.
In young sycamore, buds are produced terminally or along the shoot axis. With age some of these buds, both terminal and lateral, will also produce an inflorescence and a bud on either side of the inflorescence. The following year the shoots will fork producing two shoots with roughly a 45-90° angle. Even if one of the buds dies or if only one bud is formed, the following year's shoot will be at a 25-45° angle. Lammas shoots have not been recorded on sycamore in Ireland. By a combination of bud scars and shoot angles it is possible to estimate the percentage of buds flowering over a period of at least 10 years. In exposed areas bud death occurs regularly, but these can be easily recognized on shoots up to 5-7 years old. Because the accuracy of the flowering estimate decreases with shoot age it was decided to use only the data of the previous six years. In sycamore a good flowering year is usually followed by a poor year (see Section 2.7), and by selecting six years the estimate possibly includes three such cycles. Flowering intensity is defined as the average percentage of buds flowering over six years. These estimates were obtained by averaging the four percentages for each year and then averaging the six year's estimates. These estimates do not give any indication of the total number of inflorescences and flowers produced.
The age of the trees was estimated using the girth-age relationship obtained in Binggeli & Rushton (1985).
RESULTS
Initiation of flowering varies widely with GBH in the invasive population (Fig. 2). Some trees with GBHs up to 63cm (estimated age 54 years) have yet to flower, whereas flowering was recorded on a tree with a GBH of 29cm (27 years) with a flowering intensity of 36%. Within any 10cm GBH class a large variation in flowering intensity is observed. Following a peak in flowering intensity around a GBH of 80 to 90cm (maximum is 75% at GBH 81cm), percentage flowering decreases with size and much greater variation in flowering intensity is observed among large trees of similar size.
CHART GOES HERE
Figure 2. Percentage flowering of trees of varying GBH for the period 1983-1988. þ denotes the mean with standard of deviation of six year average flowering (1981-86) in a population of seven adult trees (with GBH between 200cm and 250cm) from the nearby Breesha plantation investigated in flowering periodicity.
The impact of flowering on tree architecture is illustrated (Fig. 3) on the photographs of two trees (S29 and S27 with GBH of 38cm and 81cm respectively). Tree S29 has yet to flower, whereas on tree S27 74% of the buds flowered over the period of 6 years (highest figure recorded in the wood). Tree S29 bears branches with clear apical dominance without forking along the main axis and has short side branches. On tree S27 the branches are convoluted and in their terminal part no clear leader is observed.
 |
 |
| Figure 3. Impact of flowering intensity on tree architecture in sycamore. A. tree S29 which has yet to flower and B. tree S27 on which 74% of the buds flowered over a period of six years. | |
These observations do not support Hall et al.'s (1978) statement that species of Acer, including sycamore, are typical example of Rauh's architectural model. Their statement is based on Rauh's (1939) paper where the architecture of a sycamore sapling is described and said to be typical of that model but no description of the tree architecture is provided and Hall‚ et al. (1978) must have assumed it to be the same. The definition of Rauh's model include the following condition: "Flowers are always lateral and without effect on the growth of the shoot system." (Hall‚ et al. 1978, p 221). Recently, Roloff (1989) noted that adult trees followed Scarrone's model where shoots are initiated below the inflorescences which occur terminally. In reality there is a gradual shift in tree architecture, which Bell (1991, p 296) described as "changes of branching development during their lifespan which in the context of models can be described as a switch from the model of Rauh to Scarrone to Leeuwenberg. The timing of these events will depend upon environmental conditions, particularly the degree of shading." In fact in open, but sheltered, conditions such changes are solely related to flowering intensity.
DISCUSSION
The data on flowering intensity in the invasive population suggest that phase change occurs between a GBH of 29 and 63cm (about 27 to 54 years of age), reaches a flowering peak at size classes of 80 to 90cm (67 to 76 years) and then decrease with increasing tree size. The age of first flowering and of maximum relative flowering intensity is greater than the usually reported 15-20 years for phase change and 25-30 years for the first good seed crop. The time difference is probably due to a combination of a/ poor soil conditions, b/ exposure, particularly to northerly and easterly winds and, c/ to the north-easterly aspect of the wood reducing direct sunlight for a greater part of the day. These factors are described in detail in Section 5.2.
The decrease in flowering intensity in trees with large GBHs is likely to be caused by similar factors and wind is likely to be more important as large sycamores emerge from the birch-rowan canopy. The large trees from the nearby Breesha plantation investigated in the flowering periodicity study showed a much greater percentage of buds flowering every year. The average flowering intensity of seven trees (GBH = 228±15.8cm, range 200-250cm) between 1981 and 1986 was 63.7±16.0% and is shown on Fig. 2. The three other sycamore populations investigated in Northern Ireland (see flowering periodicity) show that a high percentage of buds on most adult trees produce inflorescences over several flowering cycles. Old trees produce fewer buds but the percentage of these buds flowering is still high.
Competition with other tree canopies could also reduce flowering on some trees. It has often been noticed that trees in high forest stands have a late age of phase change and that their seed production is much less than that of trees in open conditions. Although sycamores in the birch-hazel-rowan wood suffer from less competition than in other forest types, the unfavourable environmental conditions appear to severely restrict flowering. All trees with high flowering intensities were located in areas which were either more sheltered from the wind or with a more developed soil cover with fewer large boulders, or a combination of the two.
Following a decade of qualitative field observations, I have observed that protogynous trees always carry a crop of seeds whenever and wherever they flower. Although no data are available on seed survival and germination rates of these trees, it is likely that in the invasive population, recruitment of young sycamores takes place without the input of seeds originating from planted specimens outside the woodland.