Shoot growth in relation to cut flower production of Banksia coccinea and Banksia menziesii (Proteaceae)

Scientia Horticulturae, 49 ( 1992 ) 323-334 Elsevier Science Publishers B.V., Amsterdam

323

Shoot growth in relation to cut flower production of Banksia coccinea and Banksia menziesii ( Proteaceae ) A.M. FuSS a'l, S.J. Pattison b, D. Aspinall a and M. Sedgley a,2 aDepartment of Horticulture, Viticulture and Oenology, Waite Agricultural Researck Institute, The University of Adelaide, P.O. Box l, Glen Osmond, S.,~. 5064, Australia bBiometry Section, Waite Agricultural Research Institute, The University of Adelaide, P.O Box !, Glen Osmond, S.A. 5064, Australia (Accepted 10 July 1991 )

ABSTRACT Fuss, A.M., Pattison, S.J., Aspinail, D. and Sedgley, M., 1992. Shoot growth in relation to cut flower production of Banksia coccinea and Banksia menziesii ( Proteaceae ). Scientia Hortic., 49: 323-334. Shoot growth in relation to the flowering of Banksia coccinea and Ban~ia menziesii was investigated to aid the development of breeding criteria and plantation management strategies for cut flower production. In both species there was a positive relationship between shoot length, shoot diameter and number of leaves, and the probability of a shoot producing an inflorescence. The majority of blooms of both species were initiated on shoots in their second year of growl:a, although 28% of B. coccinea and 9% ofB. menziesii shoots flowered in t!:eir first year. Twenty-nine percent orB. coccinea and 20% of B. menziesii shoots did not produce aI~ inflorescence with|n the 17-month observation period. There was considerable variation between plants in the number of blooms, the number of leaves per shoot, leaf size and the probability of producing laterals. These factors influence both productivity and the overall floral display of the cut flower and can be used as criteria for plant improvement through selection and breeding. Keywords: Banksia coccinea; Banksia menziesii; cut flowers; flowering; phenology; Proteaceae; shoot growth.

INTRODUCTION

Banksias are woody evergreen plants native to Australia, which produce blooms annually, after the completion of a juvenile phase (George, 1987; Lamont and Barker, 19'88). Banksia coccinea and Banksia menziesii are two IPresent address: Department of Agriculture, Baron-Hay Court, South Perth, W.A. 6151, Australia. 2Author to whom correspondence should be addressed.

© ! 992 Elsevier Science Publishers B.V. All rights reserved 0304-4238/92/$05.00

324

A.M. FUSS ET AL.

Western Australian species which are cultivated in a number of countries for their conspicuous terminal inflorescences (Elphick, 1985 ). Little is known of their cultural requirements and, because there is as yet no clonal propagation of material, the plants exhibit considerable variation in form, bloom production and colour. Banksia coccinea blooms are predominantly scarlet, but range from orange through to dark red. In Australia they are harvested between July and November, some 9-12 months after floral initiation in late spring (Fuss and Sedgley, 1990). Floral initiation in B. menziesii also occurs in late spring, subsequent floral development takes 6-8 months, and the flowering period extends from February to September, with a peak in May. The majority of B. menziesii blooms are pink, although yellow and dark-red variants also exist. Growth in both species follows Leeuwenberg's model (Halld et al., 1978). Shoot extension growth occurs from the terminal bud during each active growth season until an inflorescence is initiated in the terminal position or the terminal bud is damaged. Upon termination of extension growth new lateral shoots arise from axillary buds. This study investigates shoot growth of bushes of B. coccinea and B. men:iesii grown for cut flower production, with the aim of determining the relationship between the age and size of shoots and their likelihood of flowering. Such information will be beneficial in developing strategies for plantation management, particularly pruning and yield prediction, for selection of superior plants for future clonal propagation, and for use in breeding programmes (Fuss and Sedgley, 1991 a,b). MATERIALS AND METHODS

Plant material and ctdtivation. - Seedlings of B. coccinea and B. menziesii were planted in 1983 on a commercial cut flower plantation at Blewitt Springs, South Australia (latitude 35 ° 10' S; longitude 138 o34' E) on the eastern slope of a steep sand hill (pH 6.6). The plants were arranged in rows, with 1.8 m between plants and 3.0 m between rows for B. coccinea, and 2.3 m between plants and 3.4 m between rows for B. menziesii. All plants were subjected to routine plantation management ~ractices, including limited drip-irrigation to supplement rainfall, fertilising am~ually in autumn with Complete Mineral Mix (Top Aus':ralia Ltd.) and biannually with low phosphorus Osmocote (Sierra), and mowing to control weed :~rowth. Harvesting of blooms in previous years had served as the only form of pruning. Meteorological data for the experimental period, March 1988 to A~gust 1989, are presented in Fuss and Sedgley (1990). Measurements. - In March-April 1988, six plak~:ts of each species were selected and all shoots longer than l 0 cm on B. coccim,,q and 5 cm on B. menziesii were labelled. This represented approximately 90% of the total number of

SHOO'L GROWTH AND FLOWER PRODUCTION OF B..INKSIA

325

shoots per bush of B. coccinea and 95% of the total number of shoots per bush of B. me:rziesii. Those shoots with a single growth flush were tagged at the base of the shoot, while those more than 1 year old were tagged at the most recent trans!tion between the previous and the latest flush, as recognised by the presence of bud scales (Fi ;. l a). Shoot growth measurements were recorded at monthly intervals until August 1989 so that each species went through one com~,lete harvest. Measurements were made on all labelled shoots of B. menziesii and on every second shoot of B. coccinea. The length of each growth flush was measured to the nearest 0.5 cm, from the point of attachment of the label to either the base of the terminal bud or the developing inflorescence, or a bud scar. The length of any previous growth flushes ( 1986 and 1987 ) was measured a~ the first recording time only. The basal diameter of each flush was taken 1 cm above the base of the flush, at the point of attachment of the label, to the nearest 0.1 mm using manual calipers. At the completion of extension growth 6f the flush in both years, in July for B. coccinea and March for B. menziesii, the number of leaves per flush was recorded. In the first year only the length, including lamina and petiole, and width of the fifth leaf from the base of the flush, were recorded to the nearest 0.1 cm. The vegetative/floral status of e~ch shoot and the development of laterals was recorded throughout the experimental period. Flowering shoots were removed following anthesis. Statistical analysis.- Shoots were categorised depending on the number of growth flushes from the axillary origin of that shoot and the development status at the end of each flowering season. The categories of shoots were those which produced an inflorescence that bloomed in the first year of growth (1988), produced an inflorescence that aborted in the first year of growth ( 1988 ), produced an inflorescence that bloomed in the second year of growth ( 1988 or 1989 ), produced an inflorescence that aborted iI~ U:e second year of growth ( 1988 or 1989), or did not initiate an inflorescence within the first 2 years of growth ( 1988 or 1989). The number of shoots in each category as a proportion of the total per tree was analysed as binomial data. Th~s generated a chi-square test for differences between plants in the proportion of shoots per plant for a given shoot category. Analysis of variance (ANOVA) vas conducted to test for differences in shoot length and basal diameter due W shoot category, having first adjusted for a possible plant effect. Similar analyses were used to test for differences in shoot length and basal diameter on compar,~ble growth flushes occurring in different years to investigate the effect of season. The relationship in both years between shoot length and shoot diameter" number of leaves, and shoot length and shoot diameter; and length and width of the fifth leaf, were subjected to regression analysis, taking plant-to-plant variability into consideration. Those ~hoots that did not produce a bloom within 2 years growth were excluded from the regression analysis. The num-

326

A.M. FUSS ET AL.

b

SHOOT GROWTH AND FLOWER PRODUCTION OF B..INKSIA

327

ber of shoots producing laterals out of the total number of shoots for each shoot category and each plant was analysed as binomial data. This resulted in a chi-square test to determine if estimates of the probability of a shoot developing laterals were based simply on the total proportion of shoots producing laterals, or were affected by shoot category or plant-to-plant variability alone, or by an additive effect of shoot category and plant-to-plant variability. RESULTS

Only 28% and 9% of the shoots of B. coccinea and B. menziesii, respectively, flowered in their first year of growth, with the majority of inflorescences initiated on second,, year wood (Table 1 ). In B. menziesii a proportion of the inflorescences initiated on both first- and second-year wood aborted in the early stages of devezopment. There were significant differences between B. coccinea plants in the proportions of shoots which produced blooms in their first year of g r o ~ h (P<0.01) or in their second year of growth (P<0.01), whereas in 3. menziesii there was no significant difference between plants in the proportion of shoots which initiated inflorescences in their second year of growth. There was, however, a significant difference between trees in the initiation of blooms on first-year wood, irrespective of whether they developed through to anthesis ( P < 0.5 ) or aborted during development (P < 0.01 ). Two trees in particular contributed to this difference between trees, having a much higher proportion of shoots in these categories than the other four trees. Twenty-nine percent of the labelled shoots of B. coccinea did not produce a bloom within 2 years of growth, and the proportion of non-productive shoots was similar on each of the six bushes sampled. Shoots of both species which produced a bloom in their first year of growth were longer and thicker than [hose w.hich did not .(Table ! ). Afte.r..I.year's growth there were significant differences in both shoot length (P<0.01) and basal shoot diameter ( P < 0.01 ) between shoots which had flowered in their first year, those which would go on to flower in their second year, and those which, after a further year of growth, would not produce a bloom. There were similar significant differences in the length (P< 0.01 ) and diameter ( P < 0.01 ) of floral and non-floral shoots in t.he second year's growth. In any given year, shoots that were floral were longer and thicker than the non-floral shoots. Active growth of shoots of both species commenced in October and resuited in a rapid increase in the basal diameter of the previous year's growth, and extension growth from the terminal bud (Fig. 2). After February there Fig. 1. (a) B. coccinea shoot which flushed once in a given season showing a ring of bud scars (arrow) at the base of the flush. (b) B. coccinea shoot which flushed twice in a given season showing bud scars separated by extension growth at the base of the second flush (arrow) resulting in a region of stem devoid of leaves (I). Bar represents 5 cm.

328

A.M. F U S S ET AL.

• ,.,.,

C'-- , - , •

-t.1

II

II

'-"

,'M

II

II

II

II

o

•

I

-H

~C~ ~ 1 0

¢",1

• I

+1

+1

-t-I

-I-I

-H

+1

-H

-H

-t-I

-+ ]

-H " "

-I-I "

-H

~

O

. ,...,

o,,~ • I

•

•

o

~'~ O +1

I"-

O -t-I

+1

o,~

-H

I

I

-"

-H

-I-I

-H

O

I'-. O

,/'~ o

O i=.. O

("4

.,u

~,

I'-- ¢ ' 1 0

O

O

O

,.~ .o ~'~ -t-I ~

.~ -I-I " "

-t-I

"

-t-I--'

+1

u-~ oO {,-4 ~

O

,r~ I'...

•

~" +1 "

+1

+1 - "

+1 --' +1

+1

+1 ~

+1 ~

-j O

O

Z'

+1

..I

o.I

o

o

+1

.=_

.=_

~~ . ~~= ~ .

a. ~ .

+1

,~

O t.

+1

@ O

O O

=~ O

O .,,..

O ..N

....,

~n

,< Z

.=

~

"-

88~ ~ ~i 5 . =.

329

S H O O T G R O W T H A N D F L O W E R P R O D U C T I O N O F BANKSIA 15

a

E 10 E

._m a

~ s (10

0 60 50

A

4O

30 2.o 10

MAMJ

J

AS

1988

ON

DJ

FMAMJ

1989

J

A

MAMJ

J

A S ON

1988

DJ

FMAMJ

J

A

1989

Fig. 2. Changes in basal shoot diameter and shoot length for B. menziesii shoots which flowered in their second year of growth (a), and those which did not flower after 2 years growth (b), from monthly observations from March 1988 to August 1989. Basal diameter of first year's growth ( O ) and second year's growth ( A ) had a maximum SE of 0.5 mm for both shoot types. Maximum SE for first (solid bars) and second (shaded bars) year's shoot length was 2.1 cm.

was no further increase in shoot length for either floral or non-floral shoots, but basal diameter continued to increase. Both extension growth and shoot diameter were greater in the flowering than in the non-flowering shoots. Those shoots orB. menziesii which initiated inflorescences that subsequently aborted were intermediate between the floral and the non-floral shoots for both length of extension growth and basal diameter. Some shoots of B. coccinea produced more than one growth flush within a season (Table 2). Within any growth season it was possible to distinguish between the different flushes. At the base of the first flush for the season, there was a ring of bud scars around the stem (Fig. 1a). Subsequent flushes within that season, were characterised by bud scars separated by extension growth (Figo I b). There was a tendency for shoots which flushed twice not to flower in that season, with the highest proportions of blooms being produced on shoots after a single growth flush (Table 2). There was a positive relationship in B. coccinea between shoot length and

330

A.M. FUSS ET AL.

TABLE 2

Total number of axillaD' shoots and a breakdown of the types of growth on six trees o f B. coccinea with only one season's growth at the commencement of the experiment Number of shoots and type of growth flush in the xth year x=l 60 109

~

O

Percentage o f total number of shoots

x=2 27% 62

28%

34 5

8

.

O

>m, ,,

>

---o

15% 2%

4% 12%

28

16 5

7% o

2%

2 , >m,

1%

5___,_.. o

20/0

*, inflorescence; o, resting terminal bud produced at the end of an active growth season; >, terminal bud produced during an active growth season.

basal diameter of shoots which produced a bloom on first- or second-year wood based on the 1988 data (R2=0.75). In the following year the relationship was much weaker (R2=0.48). Data from both years showed betweenplant variation. In B. menziesii there was a poor correlation between shoot length and basal diameter in both years of the study (R2=0.48 and 0.37 respectively). There was a strong positive correlation in both years between the number of leaves on a shoot and that shoot's length and basal diameter (B. coccinea R2=0.82 and 0.83 respectively; B. menziesii R2=0.87 for both years). There was variation between plants in both years, although the influence of shoot diameter on the number of leaves in B. coccinea was less variable between plants in the first year. Plants ofboth species also acted independently with regard to the relationship between the length and width of the fifth leaf, indicating between-plant variation in leaf shape (B. coccinea R2=0.55; B. menziesii R 2 =0.81 ). The effect of season on shoot growth was considered by comparing retrospective observations of length and diameter of similar flushes on shoots which developed from axillary buds in 1986 and 1987, and flowered in their second year. There was no significant difference in the length of the first flush or the diameter of the second flush of either species. However, the second flush of the 1986 shoots of B. coccinea was significantly longer than that of the 1987 shoots ( 35.4_+ 2.4 cm and 28.5 _+ 1.0 cm respectively, P< 0.01 ). Lateral shoots of B. coccinea developed around the base of developing in-

33 i

SHOOT GROWTH AND FLOWER PRODUCTION OF BANKSIA

TABLE 3 Estimates of the probability of a shoot producing laterals for each of six plants of B. coccmea ana B. menziesii. In 1988 and ! 989, the estimates for B. coccinea were based on an additive effect of plant and shoot category, while for B. menzwsii plants in ! 988, the estimates were based on shoot category only (there was no plant-to-plant variation in 1988 ), and in 1989, on the proportion of shoots producing laterals. Those shoots which did not produce an inflorescence in 1988 were followed through to 1989as (a), (b) and (c) Probability of a shoot producing laterals (plant number)

Shoot category

B. coccinea 1988 Inflorescence in year 2 Inflorescence in year I No inflorescence in year ! 1989 (a) Inflorescence in year 2 (b) No inflorescence in 2 years B. menziesii 1988 Inflorescence in year 2 Aborted inflorescence in year 2 Inflorescence in year i Aborted inflorescence in year ! No inflorescence in year 1

(a) (b)

!

2

3

4

5

6

0.35 0.02 0.00 0.00

0.63 0.07 0.00 0.00

0.90 0.28 0.00 0.00

0.29 0.02 0.00 0.00

0.90 0.28 0.00 0.00

0.83 0.17 0.00 0.00

0.24 0.00

0.70 0.00

0.23 0.00

0.20 0.00

0.89 0.00

0.54 0.00

0.50 0. ! I 0.00

0.60 0.00 0.50

0.95 0.27 0.00

0.40 0.00 0.00

i.00 0.60 0.00

0.32

(a) (b) (c)

1989 (a) Inflorescence in year 2 (b) Aborted inflorescence in year 2 (c) No inflorescence in 2 years

0.06 0.06 0.00 0.00 0.00 0.41 0.25 0.14

*Too few observations to be included in analysis.

florescences at an acute angle to the stem. As the inflorescence increased in width during development, the rows of florets were forced to part around adjacent laterals causing permanent distortion to the symmetry of the inflorescence. Lateral shoots on B. menziesii grew out at a wide angle from the main stem and did not interfere with the structure of the developing inflorescence. In both 1988 and 1989 there was an additive effect in B. coccinea of plant and shoot category on the development of lateral shoots (Table 3). Shoots that did not produce an inflorescence did not produce laterals either. In B. menziesii there were differences in the probability of a shoot producing laterals between shoot categories in 1988 but no betw,zcr~-plant variation as there was in B. coccinea. In 1989 there was variability between plants as well as between shoot categories in the production of laterals.

332

A.M. FUSS ET AL.

DISCUSSION

Plant variability is an important consideration in selection of material for clonal propagation and for crop improvement through breeding. In B. coccinea and B. menziesii the majority of blooms were initiated on shoots in their second year of growth, although some shoots produced an inflorescence in the first year. Selection of plants with a tendency to produce shoots flowering in their first year and with few unproductive shoots would be an economic advantage. From the point of view of cut flower production, this study has shown that selection of such plants would be a realistic objective with B. menziesii but not for B. coccinea. Shoots of B. menziesii which flowered in their first year were extremely vigorous and attained a length in I year comparable with 2 year's growth of those flowering in the second year. However, none of the plants of B. coccinea observed either in the detailed study or generally in the plantation produced long-stemmed blooms on l-year-old wood, indicating low variability for this character, and this short stem length has limited market appeal. In comparison, shoots flowering in their second year attained a much longer total stem length. Therefore, selection of plants with a higher proportion of this type of shoot for B. coccinea would give greater flexibility in marketing the cut flower. The aesthetics of a cut flower are greatly influenced by its associated vegetative growth, and it is desirable to select for vegetative characteristics which enhance the appearance of the overall floral display (Wilfret, 1987). The double flushing observed in some of the shoots ofB. coccinea is similar to that which occurs in many tropical woody plant species, including avocado (Scholefield et at., 1985 ) and mango (Scholefield et al., 1986 ). This flushing pattern has not previously been reported for any Banksia species. However~ in the context of producing a high-quality cut flower it is undesirable, as an unattractive region of bare stem occurs at the base of the second growth flush. Leaf size, shape and density on a shoot were highly variable in both species. These may become important selection criteria as the market for Banksia blooms becomes more sophisticated and leaf characters assume greater importance. The development of lateral shoots also needs consideration. In both species shoots tend to develop laterals when terminal growth ceases, due to either initiation of an inflorescence or damage to the terminal bud. This development of lateral shoots around the base of the inflorescence interferes with the floral display and reduces the appeal of the bloom. In B. coccinea lateral shoots cause perrnJnant distortion to the developing inflorescence, while in B. menziesii they do not detract from the bloom to the same extent. However, they do interfere with floral appearance and must be removed when the blooms are graded after harvest. It is possible that removal of laterals may reduce the vase life of the bloom, as occurs in other Proteaceous species, due to the production of ethylene from the wounds (HarrY, 1988 ). Therefore se¢

1

SHOOT GROWTH AND FLOWER PRODUCTION OF BANKSIA

333

lection of plants with a low probability of lateral shoots developing around an inflorescence is desirable for the production of standard cut flowers. This study has shown that there is sufficient variability for all of these characters to allow rapid gains in selection and breeding. This is the first study to approach the development of management strategies by quantitative analysis of shoot growth. The propensity of a shoot to flower has been shown to be influenced by both the age of the shoot and its size. These criteria can be used by growers to assess which shoots to retain and which to remove during pruning. The absolute values will vary depending on species, genotype, and location of the plantation, but the relativity between shoots was a consistent feature of this study and can be adapted to other situations. Floral initiation in B. coccinea and B. menziesii cultivated under South Australian conditions occurs in late spring (Fuss and Sedgley, 1990). In B. menziesii there are several months between peak bloom production in May, and floral initiation and the commencement of new growth in October. However, in B. coccinea, the flowering season extends from July to November, with the peak in October, thus new growth commences and inflorescences are initiated while the bushes are flowering. For this reason, it is important that pruning of B. coccinea is conducted in conjunction with the harvesting of blooms. Even so, some late blooms may need to be sacrificed to ensure good production in the following year. Delaying pruning until the completion of flowering may result in the removal of potential flowering shoots or in lateral buds insufficiently developed to produce an inflorescence in the following season. Abortion of inflorescences in the early stages of development was a consistent feature ofB. menziesii, and represents a significant reduction in potential returns to the grower. The reason for this is not clear, but may relate to differences in environmental conditions between South Australia and the native habitat in Western Australia. The annual mean temperature at Blewitt Springs is 14.1 °C, which is less than the natural range of B. menziesii from Perth, W.A. (18.7°C) to Geraldton, W.A. (19.7°C). Abortion has been reported to occur in :/eiopea (Faragher, 1989 ), and also in Leucospermum paiersonii under a low-temperature regime ( 17 / 12 ° C ) plus long days (WaUerstein, 1989 ). Sub-optimal temperatures also caused flower bud atrophy in 'Baccara' roses (Zieslin and Halevy, 1975 ) and terminal flower bud aborti~on in Paeonia lactoflora (Evans et al., 1990). Some plant-to-plant variation was observed for this characteristic, so it may be possible to select cuItivars to suit particular environments. ACKNOWLEDGEMENTS

Thanks to B. and J. Bagshaw and A. and P. Keith, of Blewitt Springs, South Australia, for allowing access to their banksia plantations, and Andrew Dun-

334

A.M. FUSS ET AL.

bar and Jennie Groom for assistance with photography. This research was supported by the Playford Memorial Trust Horticultural Research Scholarship and the Doreen McCarthy Bursury of the Australian Federation of University Women, South Australia Inc., to A.M.F. and grants to M.S. from the Rural Credits Development Fund of the Reserve Bank and the Australian Special Rural Research Council.

REFERENCES Eiphick, P., 1985. Commercial production ofbanksias. In: B.B. Lamont and P.A. Watkins (Editors), Horticultu~'e of Australian Plants. Western Australian Department of Agriculture, Perth, pp. 86-91. Evans, R.M., Anders~a, N.O. and Wilkins, H.F., 1990. Temperature and GA3 effects on emergence and flowering of potted Paeonia lactiflora. HortScience, 25: 923-924. Faragher, J.D., 1989. Telopea. In: A.H. Halevy (Editor), CRC Handbook of Flowering, 6. CRC Press, Boca Raton, FL, pp. 593-597. Fuss, A.M. and Sedgley, M., 1990. Floral initiation and development in relation to flowering in Banksia coccinea R.Br. and B. menziesii R.Br. (Proteaceae). Aust. J. Bot., 38: 487-500. Fuss, A.M. and Sedgley, M., 1991a. The development of hybridisation techniques for Banksia menziesii for cut flower production. J. Hortic. Sci., 66: 357-365. Fuss, A.M. and Sedgley, M., 199,lb. Pollen tube growth and seed set of Banksia coccinea R.Br. (Proteaceae). Ann. Bot., in press. George, A.S., 1987. The Banksia Book. Kangaroo Press, Sydney, 240 pp. Hail~, F., Oldeman, R.A.A. and Tomlinson, P.B., 1978. Tropical Trees and Forests, an architectural analysis. Springer, Berlin, 441 pp. HarrY, J., 1988. Proteas: the propagation and production of Proteaceae. Fisher, New Zealand, 221 pp. Lamont, B.B. and Barker, M.J., 1988. Seed bank dynamics ofa serotinous, fire-sensitive Banksia species. Aust. J. Bot., 36:193-203. Scholefieid, P.B., Sedgley, M. and Alexander, D.McE., 1985. Carbohydrate cycling in relation to shoot growth, floral initiation and development and yield in the avocado. Scientia Hortic., 25:99-110. Scholefield, P.B., Oag, D.R. and Sedgley, M., 1986. The relationship between vegetative and reproductive development in the mango in northern Australia. Aust. J. Agric. Res., 37: 425433. Wallerstein, I., 1989. Sequential photoperiodic requirement for flower initiation and development of Leucospermum patersonii (Proteaceae). Israel J. Bot., 38: 25-34. Wiifret, G.J., 1987. The role of the plant breeder in the evaluation and breeding of new floriculture crops. Acta Hortic., 205:13-19. Zieslin, N. and Halevy, A.H., 1975. Flower bud atrophy in 'Baccara' roses. II. The effect of environmental factors. Scientia Hortic., 3:383-391.