TROPICS Vol. 14 (3) Issued March 15, 2005

Phenological characteristics of reproduction and seed formation in Durio zibethinus Murray

1) 2) 2) 3) Kazuharu OGAWA *, Ahmad Makmom ABDULLAH , Muhamad AWANG and Akio FURUKAWA

1) Laboratory of Forest Ecology and Physiology, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya 464-8601, Japan 2) Department of Environmental Sciences, Universiti Putra , 43400 UPM Serdang, Malaysia 3) Department of Biological Sciences, Faculty of Science, Nara Women’s University, Nara 630-8506, Japan *) Author to whom correspondence should be addressed. Tel: +81-52-789 4071, Fax: +81-52-789-5014, E-mail: [email protected]

ABSTRACT Phenological characteristics during the reproductive processes of flower budding, flowering, fruiting and seed formation were investigated in a Durio zibethinus Murray tree growing in an experimental field of Universiti Putra Malaysia (UPM) in Selangor, Peninsular Malaysia. The survival curve showed that the relative fall rate of reproductive organs was lower at the mixed stages of flower budding, flowering and fruiting than at the early fruiting stage. The fall of flower buds commenced after the flower budding was completed. However, the falls of flowers and fruits occurred within a few days after the formation of flowering and fruiting. The periods of flower budding, flowering and fruiting were 34, 19 and 28 days, respectively. The fruit weight growth was approximated by a logistic equation, where the intrinsic growth rate varied from 0.419 to 0.794 week-1. The forming seeds germinated or possessed germination ability at the fruit growth stage where the fruit weight reached the asymptote of the logistic growth curve. Total numbers of emerged flower buds, flowers and fruits over the whole reproductive period were 1629, 1467 and 1412 per tree, respectively. The percentage of fruits with germination ability was estimated to be 1.2% of the number of fruiting ones.

Key words: Durio zibethinus Murray, fruit growth, germination ability, reproductive processes, survival curve.

INTRODUCTION During the life cycle of woody , phenological behaviors are represented by sequential processes of reproduction, such as flower budding, flowering and fruiting (Harper 1977). Among the reproductive processes, the fruiting process is directly related to the processes of seed formation and germination. The seeds develop so as to possess the germination ability during the fruit growth. Clarifying whether or not forming seeds possess germination ability should provide important information on the dynamics of forest ecosystems, including natural regeneration. Phenological information on reproductive processes in tropical trees is limited, and only qualitative information, such as seasonality, is available (Longman and Jenik 1987; Whitmore 1984). Moreover, there is little information on ripening during fruit growth. To minimize measurement difficulties, a tropical tree, Durio zibethinus Murray (clone D99B) grown in the orchard was chosen. D. zibethinus is valued for its fruits, as well as its timber (Idris 1990; Subhadrabandhu et al. 1991; Smith et al. 1992; Yaacob and Subhadrabandhu 1995). Since wild fruit trees are known to be as productive as their descendants developed in orchards (Whitmore 1984), the present results on D. zibethinus grown in an orchard could be applied to the phenology of forest trees of this species. The measurement methods were designed to determine the survival curves of each reproductive process: flower budding, flowering and fruiting. After fruiting, the fruit growth stage (where the forming seeds possess the germination ability) was determined by fitting the logistic function to the fruit weight growth curve.

MATERIALS AND METHODS Sample tree This study was conducted on a 13-year-old (as of 1993) Durio zibethinus Murray tree (clone D99B) at an experimental field station of Universiti Putra Malaysia (UPM) in Selangor, Peninsular Malaysia. One tree was sampled. The tree height and stem diameter at breast height of the sample tree were 8 m and 34.1 cm, respectively.

Survival curve The reproductive processes were divided into flower budding, flowering and fruiting (Fig. 1). To determine the survival 222 Kazuharu OGAWA, Ahmad Makmom ABDULLAH, Muhamad AWANG and Akio FURUKAWA

curves during the reproductive processes, it was necessary to identify whether the attached and fallen Flower budding organs were flower buds, flowers or fruits. Then, all flower buds of the sample tree were marked and Flower buds n information on whether or not the petals fell, styles

n 1 turned brown or ovaries became large., was recorded. Fall Un-flowering Flowering Furthermore, fruits were classified as mature or Flowers immature. n-n 1 The attached and fallen numbers of flower buds, n 2 Fall flowers and fruits were recorded 27 times at 1 to 13 day Un-fruiting Fruiting intervals from September 19, 1993 to February 12, 1994. Fruits n-n 1-n 2 The girths of branch bases were measured in October

n 3 1993. There was one replication. Fall Un-maturing Maturing Fruit growth Mature fruits n-n 1-n 2 -n 3 The study tree commenced fruiting in early October n-n 1-n 2 -n 3 Fall 1993. On November 23, the lateral diameter (DS, cm) and

longitudinal diameter (DL, cm) of 35 attached fruits were Fig. 1. Schematic representation of the reproductive 15 processes of flower budding, flowering and un- labeled and measured by Vernier calipers (Type CD- flowering, and fruiting and un-fruiting, together and 30, Mitsutoyo, Japan). The measurement was with maturing and immaturing. nn and with repeated for 26, 20 and 19 attached fruits on December 4, subscript stand for the flower budding number 16 and 27, 1993, respectively and 18 attached fruits on and the fallen number of reproductive organs, respectively. January 7, 1994. The number of sample fruits decreased, because this species drops fruits continuously after fruiting. The fallen fruits were collected to measure the fruit size and check the germination ability of seeds. The dry mass (w, g -1 1.45 2 fruit ) of attached and fallen fruits was estimated from the allometric equation, w=0.105 (DS DL) (R =0.98). This equation was formulated from the observed data on 192 fruits (Ogawa et al. 1996).

Germination test We collected seeds from 18 fallen fruits of different fruit ripening stages from November 23, 1993 to February 12, 1994 and made the germination test (Table 1). After measuring the fresh weight of seeds, the seeds were sowed in fiberglass pots containing forest topsoil and stored on the rooftop of a building of UPM.

RESULTS Relationship between the number of flower buds and branch diameter The number of branches that produced flower buds was 13; the other 13 branches did not produce flower buds. The number of emerged flower buds, ranging from 0 to 371 per branch, positively correlated with branch diameter (P<0.05) (Fig. 2). One can infer from this positive correlation that flower budding depends on materials stored in branches and current photosynthates by leaves.

Survival curve Mixed process Figure 3 is a time line showing the number of attached reproductive organs (N) per tree in the processes of flower budding,

flowering and fruiting on the semi-log coordinates. The absolute value of gradient of the straight line after the complement

∇ ∇

∇ 1 N∇ of flower budding stands for the relative fall rate RFR＝－― ― during a given time interval (t ) (Ogawa et al. （ Nt ） 1995). While RFR was lower at the mixed stage of flower buds, flowers and fruits than at the early fruit stage, RFR became Phenology and seed formation in Durio zibethinus 223

Table 1. Fruit dry weight at the fallen time, the number and fresh weight of sample seeds used for germination test, and germination ratio during the period of November 23, 1993 to February 12, 1994. Fruit No. Fallen time Fruit weight Sample seed Germination ratio Number Fresh weight per seed [g d. wt] [g f. wt] [%]

1 Nov. 23-Dec. 2 62.5 - - 0 2 -"- 19.6 - - 0 3 -"- 31.0 - - 0 4 -"- 20.2 - - 0 5 Dec. 2-Dec. 4 10.4 - - 0 6 -"- 11.0 - - 0 7 -"- 21.8 - - 0 8 Dec. 4-Dec. 13 127.6 - - 0 9 -"- 19.5 - - 0 10 -"- 127.6 - - 0 11 Dec. 16-Dec. 27 43.8 - - 0 12 -"- 45.5 - - 0 13 -"- 74.6 - - 0 14 Jan. 19 594.6 6 27.7 100 15 Feb. 1 386.1 7 20.1 100 16 Feb. 2 571.4 6 26.6 100 17 Feb. 12 303.3 4 21.2 100 18 Feb. 12 439.6 3 28.6 100

400

350

300

250

200

150

100 Number of flower buds per branch

50

0 0246810121416 Branch diameter [cm] Fig. 2. Relationship between the emerged number of flower buds per branch and the diameter of branch base. The correlation coefficient is 0.467 (P <0.05). 224 Kazuharu OGAWA, Ahmad Makmom ABDULLAH, Muhamad AWANG and Akio FURUKAWA

10000 Fig. 3. Survival curves in the mixed process of flower budding, flowering and fruiting on

N the semi-log coordinates. ○, flower bud; □, flower bud, flower and fruit; ◇, flower and fruit; △, fruit.

1000

100 Number of reproductive organs per tree

10 Sept. Oct. Nov. Dec. Jan. Feb. Month low at the later fruit stage. Separated process Figure 4 shows a time line depicting the number of reproductive organs per tree in the separate processes of flower budding, flowering and fruiting on the semi-log coordinates. While the survival curves in the flower budding and fruiting processes showed a peak, the survival curve in the flowering process fluctuated. From the emergence curves, the periods of flower budding, flowering and fruiting were determined to be 34, 19 and 28 days, respectively. The fall and emergence curves indicate that the fall of flower buds commenced after flower budding was complete. On the other hand, the fall of flowers and fruits occurred within a few days after flowering and fruiting. Most of the fall occurred after fruiting, because the fallen amount was much higher after fruiting than after flower budding or flowering. The mature fruits (of which there was enough to be harvested) began to fall from January 17.

Total number of reproductive organs over the whole growing season As summarized in Table 2, total emerged flower buds, flowers and fruits over the whole measurement period were 1629, 1647 and 1412 per tree, respectively. The ratios of flowering and fruiting numbers to flower budding number were 90.1 and 86.7%, and the ratio of fruiting number to flowering number was 96.3%. The mature fruits were estimated to be 1.2% of the flowering or fruiting number.

Germination test Seed formation was observed in the fallen fruits after November 23. From December 16 to December 27, the appearance of seeds of fallen fruits developed well. However, it was judged that the seeds did not possess the ability to germinate, because the inside parts of the seeds were not solid, rather they were liquid or jellied. However, germination was observed in all seeds of fallen fruits collected from January 19 to February 12.

Growth curve of fruit weight The dry weight growth curve of a fallen fruit (w, g fruit-1) whose seeds had all germinated, is illustrated in Fig. 5. The growth curve was approximated by a logistic formula, W w＝― 1+kexp(-λt) (1) where t is the time in weeks from the first measurement made on November 23, W is the asymptote of w,λis the intrinsic growth rate, and k is a constant. The values of W,λand k ranged from 324.6 to 650.0 g fruit-1, 0.419 to 0.794 week-1 and 5.04 to 38.60, respectively (Table 3). Using Eq. (1), values of w/W of the fallen fruits were estimated and were from 96.2 to 99.8% at the fallen time (Table 3). Phenology and seed formation in Durio zibethinus 225

Flower budding

1000

100

10 Flowering

1000

100

10

Number of reproductive organs per tree Fruiting

1000

100

10

1 Sept. Oct. Nov. Dec. Jan. Feb. Month Fig. 4. Time line showing the number of reproductive organs per tree in during the processes of flower budding, flowering and fruiting on the semi-log coordinates. ○, survival curve indicating the number of attached organs; □, emergence curve indicating the cumulative number of emerged organs; △, fall curve indicating the cumulative number of fallen organs (cf. Kikuzawa 1982, 1983). The fruits mature enough to be harvested, began to fall from the time indicated by the arrow.

Table 2. Total numbers per tree of emerged flower buds, flowers and fruits, and percentages of flowering and fruiting over the entire measurement period of September 19, 1993 to February 12, 1994. Flower buds Flowers Fruits Total Mature Immature No. of emerged organs 1629 1467 1412 17 1395 Percentage# 100 90.1 86.7 1 85.7 $ 100 96.3 1.2 95.1 * 100 1.2 98.8

No. of fallen organs 162 55 1400 5 1383

# To the number of emerged flower buds. $ To the number of emerged flowers. * To the number of initiated fruits. 226 Kazuharu OGAWA, Ahmad Makmom ABDULLAH, Muhamad AWANG and Akio FURUKAWA

800 Fig. 5. Growth curve of dry weight, w , of a fallen fruit (fruit No. 16), in which all seeds germinated. t stands for time in weeks after the first measurement on November 600 23, 1993. The curve was approximated

] using Eq. (1). 1 -

400 [g d.wt fruit

w

200

0 0 5 10 15 t [weeks]

Table 3. Coefficients of logistic growth curve of Eq. (1) in the fallen fruits where all seeds germinated. R 2 denotes the coefficient of determination in Eq. (1). w/W in Eq. (2) stands for the ratio of fruit dry weight at the fallen time to asymptotic fruit dry weight in Eq. (1). Fruit No. Fallen date W k λ R2 w/W [g d. wt fruit-1] [week-1] [%]

14 Jan. 19, 1994 650.0 5.04 0.794 1.00 99.8 15 Feb. 1 436.3 15.73 0.671 1.00 99.4 16 Feb. 2 612.1 21.19 0.558 1.00 97.2 17 Feb. 12 324.6 10.41 0.419 0.99 96.2 18 Feb. 12 474.0 38.60 0.720 1.00 99.7

DISCUSSION Relative growth rate of fruit As summarized by Stephenson (1981), fruit weight growth is expressed by single or double sigmoidal curves. Since the fruit weight growth curve of D. zibethinus can be fitted to a logistic function of Eq. (1) (Fig. 5), the present fruit weight growth belongs in the sigmoidal curve category (Subhadrabandhu et al. 1991). The intrinsic growth rate (λ) in Eq. (1) can be regarded as the maximum relative growth rate (RGR). Theλvalues of 0.419 to 0.794 week-1 in the present species were higher than the maximum RGRs in other species: 0.303 week-1 in Pinus sylvestris strobili from sub arctic areas (Linder and Troeng 1981) and 0.342 week-1 in Cinnamomum camphora fruits from temperate areas (Ogawa and Takano 1997). Since there is no thermal seasonality in the humid tropics, it is thought that fruit, as well as the trees themselves, grow vigorously in tropical areas.

Flowering period As reviewed by Longman and Jenik (1987), individual flowers of tropical trees last for days or only hours. The duration of flowering was less than a few days for individual D. zibethinus flowers, judging from the emergence and fall curves of flowers (Fig. 4). The flowering period of individual trees, which differs among tropical species, ranges from a week or so in the gregarious species to 3-4 months in other species (Longman and Jenik 1987). Thus the present flowering period of 19 days in D. zibethinus is relatively short compared with other tropical woody species. According to Kikuzawa and Mizui (1990) and Mizui and Kikuzawa (1991), the differences in the flowering periods of individual trees are related to flowering patterns such as successive and simultaneous openings. From the present results of flowering periods of individual trees and flowers, it is clear that D. zibethinus flowers simultaneously during the short Phenology and seed formation in Durio zibethinus 227 flowering period of an individual tree.

Flowering and fruiting efficiency A large proportion of flower buds may abort and abscise before opening in spp. or Triplochiton scleroxylon in the tropical areas (Longman and Jenik 1987). On the contrary, most of the flower buds (about 90%) changed to flowers in D. zibethinus. The flowering percentage of flower buds in D.zibethinus was similar to: Aesculus turbinata in the cool-temperate area, 93% (Ogawa and Saito 1993) and Cornus sanguinea in the temperate area, 82% (Guitian et al. 1996). According to Stephenson (1981), most fruits abscise or shed at the early fruit growth stage in species showing sigmoidal fruit growth. Judging from the survival curves (Fig. 4), 98% of juvenile fruits of D. zibethinus shed at the early stage of fruit growth until the beginning of November after fruiting. Therefore, based on the fruit weight curve as proposed by Stephenson (1981), it was found that fruit abscission coincided with fruit abscission

Germination ability Seeds investigated in the present study showed the germination ability at the fruit growth stage where the fruit dry weight reached the asymptote in Eq. (2). Fruits with seeds with germination ability were regarded as fruits mature enough to be harvested. Therefore, it is estimated that the percentage of fruits with seeds with germination ability was about 1.2% of the number of initiated fruits in D. zibethinus (Table 2). Oliveira and Silva (1993) conducted a germination test under field conditions using seeds with well developed cotyledons and embryos in the mature fruits of Kielmeyera coriacea and K. speciosa in a savanna. According to their germination test, the germination probability of the seeds was 40% in K. coriacea and 9% in K. speciosa. The germination ability of D. zibethinus was 100% at the fruit growth stage where the fruit weight attained to asymptote of the logistic growth curve. Therefore, the germination probability of seeds was much higher in D. zibethinus than in K. coriacea and K. speciosa.

ACKNOWLEDGEMENTS We thank Professor A. Hagihara of the University of the Ryukyus, for his helpful advice, Mr. J. Shamsuddin of the Universiti Putra Malaysia (UPM), for generously supporting our research, and the staff of the experimental field station at UPM for access to their facilities. This work is a part of the Malaysia-Japan joint research project between Forest Research Institute of Malaysia (FRIM), UPM and National Institute for Environmental Studies (NIES), Japan. This work was supported in part by a Global Environmental Research Program Grant (No. E-4) from the Environmental Agency, Japan.

REFERENCES Bakker, J. P., Poschlod, P.., Strykstra, R. J., Bekker, R. M. & Thompson, K. 1996. Seed banks and seed dispersal: important topics in restoration ecology. Acta Botanica Neerlandica 45: 461-490. Guitian, J., Guitian, P. & Navarro, L. 1996. Fruit set, fruit production, and fruiting strategy in Cornus sanguinea (Cornaceae). American Journal of Botany 83: 744-748. Harper, J. L. 1977. Population biology of . Academic Press, London. Idris, M. Z. A. 1990. Cultivation of tropical fruits. Hi-tech Enterprise, Kuala Lumpur. Kikuzawa, K. 1982. Leaf survival and evolution in Betulaceae. Annals of Botany 50: 345-353. Kikuzawa, K. 1983. Leaf survival of woody plants in deciduous broad-leaved forests. 1. Tall trees. Canadian Journal of Botany 61: 2133-2139. Kikuzawa, K. & Mizui, N. 1990. Flowering and fruiting phenology of Magnolia hypoleuca. Species Biology 5: 255-261. Linder, S. & Troeng, E. 1981. The seasonal course of respiration and photosynthesis in strobili of Scots pine. Forest Science 27: 267-276. Longman, K. A. & Jenik, J. 1987. Tropical forest and its environment. 2nd Edn. Longman Scientific & Technical, England. Mizui, N. & Kikuzawa, K. 1991. Proximate limitations to fruit and seed set in Phellodendron amurense var. sachalinense. Plant Species Biology 6: 39-46. Ogawa, K., Furukawa, A., Hagihara, A., Abdullah, A. M. & Awang, M. 1995. Morphological and phenological characteristics of leaf development of Durio zibethinus Murray (Bombacaceae). Journal of Plant Research 108: 511-515. Ogawa, K., Furukawa, A., Abdullah, A. M. & Awang, M. 1996. Analysis of translocatory balance in (Durio zibethinus) 228 Kazuharu OGAWA, Ahmad Makmom ABDULLAH, Muhamad AWANG and Akio FURUKAWA

fruit. Tree Physiology 16: 315-318.

Ogawa, K. & Takano, Y. 1997. Seasonal courses of CO2 exchange and carbon balance in fruits of Cinnamomum camphora. Tree Physiology 17: 415-420. Ogawa, T. & Saito, H. 1993. Cost of seed production in an Aesculus turbinata forest in 1985. Bulletin of the Kyoto Prefectural University Forests 37: 1-9. Oliveira, P. E. & Silva, J. C. S. 1993. Reproductive biology of two species of Kielmeyera (Guttiferae) in the cerrados of central Brazil. Journal of Tropical Ecology 9: 67-79. Smith, N. J. H., Williams, J. T., Plucknett, D. L. & Talbot, J. P. 1992. Tropical forests and their crops. Cornell University Press, Ithaca. Stephenson, A. G. 1981. Flower and fruit abortion: proximate causes and ultimate functions. Annual Review of Ecology and Systematics 12: 253-279. Subhandrabandhu, S., Schneemann, J. M. P. & Verheij, E. W. M. 1991. Durio zibethinus Murray. In: Edible fruits and nuts. (eds. E. W. M. Verheij & R. E. Coronel) pp. 156-161, Plant Resources of South-East Asia 2, Pudoc. Whitmore, T. C. 1984. Tropical rain forests of the Far East. 2nd Edn. Clarendon Press, Oxford. Yaacob, O. & Subhadrabandhu, S. 1995. The production of economic fruits in South-East Asia. Oxford University Press, Oxford. Received 13rd Mar. 2004 Accepted 15th Dec. 2004