Summary In Larix, reproductive buds most often occur ter-minally on short shoots, but they can also differentiate in lateral positions on long shoots. The phenology of long and short shoots differs considerably, with short shoots breaking bud and expanding about 5 weeks before the start of long shoot exten-sion. Foliar sprays of GA4/7 were applied to 160 branches on 10 greenhouse-grown Larix laricina (Du Roi) K. Koch grafts either before (early) or after (late) the start of long shoot extension, or during both periods, to test whether the timing of GA4/7 application affects flowering on short and long shoots. All three treatments induced flowering on both long and short shoots. The early GA4/7 treatment led to slightly, but not sig-nificantly, more flowering on short shoots than on long shoots, whereas the late GA4/7 treatment resulted in increased flower-ing on both types of shoots, but primarily on long shoots. Application of GA4/7 during both periods did not result in increased flowering over the early or late treatment alone. Based on the fact that gibberellins are metabolized rapidly in conifers and our finding that GA4/7 applied before shoot elon-gation, when the bud primordia were at a very early stage of development (detectable bud differentiation only occurred sev-eral weeks later), induced flowering on long shoots, we con-clude that the early GA4/7 treatment did not affect differentiation as it was occurring, but somehow predisposed the bud primordia to differentiate reproductively.
Keywords: bud primordia, differentiation, flower induction.
Introduction
Gibberellin A4/7 is commonly used to promote flowering in trees of the Pinaceae for research purposes and less commonly for commercial seed production. The mechanism by which GA4/7 promotes flowering is not known. However, the impor-tance of the timing of GA4/7 application, as well as the timing of adjunct treatments, has been demonstrated (Greenwood 1982, Ross 1985, 1988, Owens and Colangeli 1989, Eysteins-son and Greenwood 1990, Ho 1991). Pharis et al. (1987) proposed that flowering only occurs when endogenous con-centrations of the less polar gibberellins (including GA4 and GA7) exceed vegetative demand during bud differentiation. However, in some cases, application of GA4/7 before detectable
differentiation of bud primordia is more effective in promoting flowering than application close to or during the period of differentiation (Owens and Colangeli 1989, Eysteinsson and Greenwood 1990, Ho 1991).
In the genus Larix, reproductive buds most often occur in the terminal position on short (dwarf) shoots, but they can also occur laterally on long shoots. The phenology of the two types of shoots is very different, with short shoots breaking bud about 5 weeks before the start of long shoot extension and also setting bud much earlier than long shoots. Powell et al. (1984) reported flowering on long shoots of tamarack (Larix laricina (Du Roi) K. Koch), and we have observed female flowering on long shoots of Japanese larch (Larix leptolepis Gord.) and European larch (Larix decidua Mill.). Flowering on young, indoor-grown tamarack grafts in response to GA4/7 occurred primarily on long shoots when GA4/7 was applied after the start of long-shoot extension (Eysteinsson and Greenwood 1993), suggesting that timing of flower induction treatment influ-ences the pattern and frequency of flowering.
The experiment described here was designed to test whether early applications of GA4/7, i.e., before the start of long shoot extension, would result in increased flowering by short shoots. This question is of practical importance to seed production in larch because the timing of flower induction treatments could target different types of shoots. Thus, optimum cone produc-tion on both types of shoots might require inducproduc-tion treatments at different times.
Materials and methods
Sixteen first-order branches on each of 10 indoor-grown tama-rack grafts were selected for uniformity of size from through-out the crown. The grafts, which represented five clones and two ramets per clone, were beginning their fourth growing season from grafting. Ortet age ranged from 28 to 56 years.
The branches were randomly divided into four treatment groups each of 40 branches, with four branches per treatment per graft. One group received weekly GA4/7 foliar spray appli-cations for 5 weeks starting March 13, 1990, about 1 week after short shoot bud burst, and ending April 10 (early treatment). Another group received weekly GA4/7 foliar spray applications for 5 weeks starting April 17, at about the start of long shoot
Flowering on long and short shoots of
Larix laricina
in response to
differential timing of GA
4/7
applications
THROSTUR EYSTEINSSON
1and MICHAEL S. GREENWOOD
2 1 Iceland Forestry Service, P.O. Box 98, 700 Egilsstadir, Iceland2 Forest Biology Department, University of Maine, Orono, ME 04469, USA
Received May 31, 1994
Tree Physiology 15, 467--469
extension, and ending May 10 (late treatment). The third group received 10 weekly GA4/7 foliar sprays starting March 13 and ending May 10 (spanning both periods). The fourth group received control sprays weekly spanning both periods.
The GA4/7 was applied as an aqueous foliar spray of 200 mg l−1 in 5% ethanol with 0.2 ml l−1 Aromox C/12w as a surfac-tant. Control branches were sprayed with the same solution minus the GA4/7. The foliar sprays were applied to the entire branch at each application, which means that only short shoot foliage was sprayed during the early period, whereas both short shoot- and expanding long-shoot foliage were sprayed during the late period.
Buds were counted in spring 1991, when pollen cone, seed cone and vegetative buds could easily be distinguished. Bud counts were analyzed by ANOVA and Duncan’s multiple range test to separate means.
Results
Both the early and late GA4/7 applications resulted in increased female flowering on short shoots (P < 0.001). Male flowering on short shoots was only significantly increased by the GA4/7 treatment that spanned both periods (P < 0.027). The early GA4/7 applications (before the start of long shoot extension), led to slightly, but not significantly, more male and female flowering on short shoots than the late applications (after the start of long shoot extension) (Table 1).
Male flowering on long shoots was significantly increased by the late GA4/7 treatment and the treatment spanning both periods (P < 0.001), but not by the early treatment. Female flowering on long shoots was increased 6-fold by the early GA4/7 treatment and 14-fold by the late treatment (P < 0.001). Differences between the late GA4/7 treatment and the GA4/7 treatment spanning both periods were nonsignificant for both sexes at both shoot positions.
At the end of the 1990 growing season, the current-year long shoots comprised an average 63% of the length of each branch, and the short shoots comprised 37%. Shoot length was a good indicator of the number of potential flowering sites because the number of buds per unit of branch length (1.4 ± 0.1 buds cm−1) was strongly conserved in the greenhouse-grown trees. Across all treatments, long shoots bore 60 and 73% of the total number of pollen and seed cones, respectively. Roughly the same proportion of buds differentiated reproductively on the two
types of shoots, about 11% of short shoot buds and 14% of long shoot buds.
Average shoot length (including laterals) at the end of the 1990 growing season was 92 cm. Application of GA4/7 did not affect shoot length.
Discussion
The effects of differential timing of GA4/7 applications on flowering by short shoots were inconclusive. The optimum time of GA4/7 application for stimulating flowering on long shoots was during early shoot elongation. The late treatment may have been better than the early treatment because of the increasing amount of expanding foliage in proximity to the primordia affected, leading to greater uptake of GA4/7.
The treatment spanning both time periods (lasting 10 weeks) did not result in significantly more flowering than the early or late treatments, indicating that five weekly GA4/7 applications were sufficient to maximize flowering. Previous work has indicated that as few as two weekly GA4/7 foliar spray applica-tions can significantly induce flowering (Eysteinsson et al. 1993). We conclude, therefore, that repeated applications are not required to induce individual primordia to differentiate reproductively. However, repeated applications are more effec-tive than a single application because of clonal differences and phenological differences between branches within a tree, and because of the increased chance that sufficient GA4/7 will reach the maximum number of target buds.
Early GA4/7 applications enhanced female flowering on both long and short shoots. Lateral bud primordia are visible soon after a long shoot starts to elongate (Frampton 1960), but may be detectable before elongation (John N. Owens, personal communication). Detectable differentiation of reproductive buds occurs in short shoots of western larch (Larix occidentalis Nutt.) and, we can assume, in other larch species as well at about the end of long shoot elongation (Owens and Molder 1979). We can also assume that lateral buds on long shoots do not differentiate earlier than short-shoot buds. Thus, GA4/7 was effective at a very early stage of bud development, as has been found for western hemlock (Tsuga heterophylla (Raf.) Sarg.) where the best flowering results were obtained when GA4/7 was applied before bud burst (Owens and Colangeli 1989).
For many species, GA4/7 appears to promote flowering only when applied before visible bud differentiation (Ross 1985, 1988, Owens and Colangeli 1989, Eysteinsson and Green-wood 1990, Ho 1991). There is also evidence that GA4/7 is rapidly metabolized in conifers (Dunberg et al. 1983, Moritz et al. 1989, 1990). Taken together, these observations suggest that the signal prompting buds to differentiate reproductively can occur when bud primordia consist of relatively few cells and is maintained through many cell divisions until bud differ-entiation takes place. The nature of this signal is not known, but one possibility is that GA4/7 affects methylation or de-methylation of genes that control cone bud development.
Table 1. Numbers of pollen and seed cones produced terminally on short shoots (SS) and laterally on long shoots (LS) of indoor-grown tamarack grafts in response to GA4/7 applied before (early) or after (late) the start of long shoot extension, or during both periods. Treatment Pollen cones per branch Seed cones per branch
References
Dunberg, A., G. Malmberg, T. Sassa and R.P. Pharis. 1983. Metabo-lism of tritiated gibberellins A4 and A9 in Norway spruce, Picea
abies (L.) Karst. Plant Physiol. 71:257--262.
Eysteinsson, T. and M.S. Greenwood. 1990. Promotion of flowering in young Larix laricina grafts by gibberellin A4/7 and root pruning.
Can. J. For. Res. 20:1448--1452.
Eysteinsson, T. and M.S. Greenwood. 1993. Effects of maturation and gibberellin A4/7 on flowering and branching characteristics of Larix
laricina. Can. J. For. Res. 23:14--20.
Eysteinsson, T., M.S. Greenwood and J.E. Webber. 1993. Manage-ment of a prototype indoor orchard for accelerated breeding of larch. Maine Agric. Exp. Stn. Misc. Rep. 377, 18 p.
Frampton, C.V. 1960. Some aspects of the developmental anatomy of the long shoot in Larix decidua Mill., with particular reference to seasonal periodicity. New Phytol. 59:175--191.
Greenwood, M.S. 1982. Rate, timing and mode of gibberellin applica-tion for female strobilus producapplica-tion by grafted loblolly pine. Can. J. For. Res. 12:998--1002.
Ho, R.H. 1991. Promotion of cone production in potted black spruce grafts, using gibberellins, heat-treatment and root-pruning. For. Ecol. Manage. 40:261--269.
Moritz, T., J.J. Philipson and P.C. Odén. 1989. Metabolism of tritiated and deuterated gibberellins A1, A4 and A9 in Sitka spruce (Picea
sitchensis) shoots during the period of cone-bud differentiation. Physiol. Plant. 77:39--45.
Moritz, T., J.J. Philipson and P.C. Odén. 1990. Quantitation of gib-berellins A1, A3, A4, A9 and an A9-conjugate in good- and
poor-flowering clones of Sitka spruce (Picea sitchensis) during the period of flower-bud differentiation. Planta 181:538--542. Owens, J.N. and M. Molder. 1979. Bud development in Larix
occiden-talis. II. Cone differentiation and early development. Can. J. Bot. 57:1557--1572.
Owens, J.N. and A.M. Colangeli. 1989. Promotion of flowering in western hemlock by gibberellin A4/7 applied at different stages of
bud and shoot development. Can. J. For. Res. 19:1051--1058. Pharis, R.P., J.E. Webber and S.D. Ross. 1987. The promotion of
flowering in forest trees by gibberellin A4/7 and cultural treatments:
a review of possible mechanisms. For. Ecol. Manage. 19:65--84. Powell, G.R., K.J. Tosh and W.R. Remphrey. 1984. Occurrence and
distribution of cones borne laterally on long shoots of Larix laricina. Can. J. Bot. 62:771--777.
Ross, S.D. 1985. Promotion of flowering in potted Picea engelmannii
(Perry) grafts: effects of heat, drought, gibberellin A4/7, and their
timing. Can. J. For. Res. 15:6178--624.
Ross, S.D. 1988. Effects of temperature, drought, and gibberellin A4/7,
and timing of treatment on flowering in potted Picea engelmannii
and Picea glauca grafts. Can. J. For. Res. 18:163--171.
