TEMPORAL AND SPATIAL PATTERNS OF MASS

Transcription

1 American Journal of Botany 90(7): TEMPORAL AND SPATIAL PATTERNS OF MASS FLOWERINGS ON THE MALAY PENINSULA 1 SHINYA NUMATA, 2,6 MASATOSHI YASUDA, 3 TOSHINORI OKUDA, 2 NAOKI KACHI, 4 AND NUR SUPANDI MD. NOOR 5 2 National Institute for Environmental Studies, 16 2 Onogawa, Tsukuba, Ibaraki , Japan; 3 Forestry and Forest Products Research Institute, Matsunosato 1, Tsukuba, Ibaraki , Japan; 4 Department of Biological Science, Graduate School of Science, Tokyo Metropolitan University, Hachiouji, Tokyo , Japan; and 5 Forest Research Institute Malaysia, Kuala Lumpur 60000, Malaysia We propose a hypothesis to explain the temporal and spatial patterns of mass flowerings in dipterocarp tree species on the Malay Peninsula. The literature on these mass flowerings reveals that during at least 11 flowerings occurred at irregular intervals of 1 6 yr in a lowland rain forest. Five of them were typical mass flowerings a high density of flowering trees and the characteristic sequential flowering of Shorea species. The 11 flowerings were classified into two flowering times: spring and autumn. There is evidence that low temperature and drought triggered the flowerings. Therefore, the seasonality of mass flowerings is characterized by the annual patterns of rainfall and low temperature. In addition, changes in El Niño Southern Oscillation (ENSO) may play important roles in determining the supra-annual occurrence of mass flowerings. Flowering surveys on the Malay Peninsula implied that regions with spring or autumn mass flowerings corresponded geographically to those regions that had one cool season (December February) or two (December February and June August), respectively. This finding anticipates the seasonal pattern and geographical distribution of mass flowerings on the Malay Peninsula. Key words: Shorea. aseasonal tropics; Dipterocarpaceae; drought; El Niño La Niña; low temperature; Malay Peninsula; mass flowering; Trees of the family Dipterocarpaceae are widely distributed in lowland rain forests of the aseasonal Southeast Asian tropics (Symington, 1943; Ashton, 1982). Even though this region has little seasonal weather variation, dipterocarps have a unique rhythm of reproductive phenology known as mass flowering or general flowering, followed by mast fruiting (Ashton et al., 1988; Sakai, 2002). Mass flowering represents supra-annual synchronization of flowering at irregular intervals of 2 10 yr. It results in a massive number of fruits in the forests during a period of 5 or 6 mo. From the many ecological and evolutionary studies on this mass flowering, several hypotheses have concerned the adaptive significance of mass flowering, including the pollination efficiency hypothesis (Norton and Kelly, 1988) and the predator satiation hypothesis (Janzen, 1974; Numata et al., 1999; Curran and Leighton, 2000; Curran and Webb, 2000). Mass flowering, however, remains notorious for its spatial and temporal unpredictability (Burgess, 1972; Yap and Chan, 1990). Burgess (1972) analyzed seasonal patterns of mass flowering from reports by the Malaysian Forestry Department spanning 11 yr. He found a distinct peak in the second quarter of the year and a smaller peak in the fourth in the central and southern parts of the Malay Peninsula, but no seasonal pattern in the northern part. Ashton et al. (1988) also argued in favor 1 Manuscript received 8 November 2002; revision accepted 30 January The authors thank the Malaysian Meteorological Service for providing the meteorological data. We thank Drs. S. J. Wright, P. S. Ashton, J. Matsumoto, C. Miyazaki, and anonymous reviewers for valuable comments and suggestions. This study was financially supported by fellowships from the Japan Society for the Promotion of Science, the Joint Research Project of the Forest Research Institute Malaysia, the Universiti Putra Malaysia, and the National Institute for Environmental Studies under the Global Environmental Research Program supported by the Ministry of the Environment of Japan, Grant No. E numata.shinya@nies.go.jp of a regional pattern of rainfall and timing of mass flowering on the basis of herbarium specimens in western Malesia, excluding the Philippines. He determined the predominant flowering months of dipterocarps in each part of the region, but found no clear correlation between flowering intensity and annual rainfall or mean temperature pattern. Despite the lack of a clear relationship between flowering and climate patterns, shared external stimuli must be the key to explaining the synchronization of flowering and fruiting within and among forest communities (Whitmore, 1984). There is evidence that some meteorological cues, including prolonged drought (Burgess, 1972; Medway, 1972; Appanah, 1985), strong solar radiation (Ng, 1977), tree nutritional status (Isagi et al., 1997), and abnormal temperature (Ashton et al., 1988; Yasuda et al., 1999), trigger floral induction in dominant canopy trees. Of these possible cues, a drop in night air temperature during cloudless weather is currently accepted as the most plausible cue for supra-annual synchronization of flowering on the Malay Peninsula (Ashton et al., 1988; Yasuda et al., 1999). Some studies have demonstrated that mass flowerings occur approximately 2 mo after several nights of low air temperatures in both western and eastern Malaysia (Ashton et al., 1988; Sakai et al., 1999; Yasuda et al., 1999). Yasuda et al. (1999) revealed that mass-flowering forests spatially coincided with meteorological stations that had recorded such low temperatures during droughts on the Malay Peninsula. Thus, low minimum temperatures during drought may be the key to understanding temporal and spatial patterns of mass flowerings on the Malay Peninsula. An understanding of temporal and spatial patterns of mass flowerings would provide important information on forest dynamics and help in forest management, because mass flowering plays a central role in the regeneration of many canopy trees in Southeast Asia. Such knowledge would also help in the recovery of disturbed rain forests where many species of

2 1026 AMERICAN JOURNAL OF BOTANY [Vol. 90 TABLE 1. Studies that have described flowerings of dipterocarp species in Pasoh Forest Reserve, peninsular Malaysia. The beginning of flowering was determined from the flowering of Shorea macroptera Dyer, which flowers first among Shorea species during a mass flowering. Year a Beginning of flowering February October August September April October August Mass-flowering type Magnitude of mass flowering Mass (ca. 55%) Mass (ca. 65%) Sporadic (ca. 15%) Mass (43%) Sporadic (13%) Sporadic (23%) Mass (unknown) Unknown Mass (unknown) Sporadic ( 10%) Sporadic (24%) Mass (55%) a For the years 1984, 1988, and there is no available information, but no mass flowering was reported. Reference Appanah, 1985 Ashton et al., 1988 Yap and Chan, 1990 Ashton et al., 1988 Yap and Chan, 1990 Yap and Chan, 1990 Ashton et al., 1988 Toy and Marshall, 1992 Toy and Marshall, 1992 Toy and Marshall, 1992 LaFrankie and Chan, 1991 Thomas, 1993 Yasuda et al., 1999 Numata et al., 1999 S. Numata, unpublished data S. Numata, unpublished data S. Numata, unpublished data tropical canopy trees have short-lived, recalcitrant seeds. We examined the temporal pattern of mass flowerings in dipterocarp tree species by analyzing flowering and meteorological data in literature on a lowland rainforest. We then tried to determine the intra- and interannual patterns of mass flowering and any meteorological cues. Finally, to extend the spatial and temporal patterns of mass flowerings from the lowland rain forest to the Malay Peninsula, we determined geographical distribution of mass flowerings by flowering surveys and collected meteorological data on the peninsula. We analyzed weather patterns on the peninsula and used them to classify flowering patterns. MATERIALS AND METHODS Flowering information in a lowland rain forest The main study site was the Pasoh Forest Reserve, located 70 km from Kuala Lumpur (latitude 2 59 N, longitude E, altitude m). The reserve is a typical lowland rain forest in peninsular Malaysia (Manokaran and Swaine, 1994). Because biological studies have been conducted here under international research projects since the 1970s, much information on forest ecosystems has been gathered here. The total area of the reserve is 2450 ha. The reserve is bordered on the east, south, and west by oil palm plantations and on the north by virgin hill dipterocarp forest. The core area of the reserve (approximately 600 ha), where the study plots were located, is generally homogeneous in topography and community structure, with no evidence of major human disturbance. The forest contains stands at various stages of maturity from canopy gaps to climax forests topped by emergent trees with heights of 50 to 60 m. The occurrence of flowerings was determined from studies conducted at the Pasoh Forest Reserve (listed in Table 1). Combining eight reports and our observations from 1996 to 2002, we determined the occurrence of mass flowerings from 1980 to We also determined the beginning of flowerings and the magnitudes of mass flowerings from descriptions in the reports. The beginning of flowering was determined from the flowering of Shorea macroptera Dyer, which flowers first among Shorea species during a mass flowering. The magnitude of a mass flowering was estimated from the proportion of trees flowering when that information was given. Flowering survey on the Malay Peninsula We surveyed the fruiting of dipterocarp trees in June July 1996 and November To determine the geographical distribution of mass flowerings, we surveyed fruiting in forest reserves, protected forests, and old secondary forests, from lowlands to hills accessible by vehicles (about 60 locations each year), and established an observation area of about 5 ha in a forest. Because dipterocarp species flower sequentially but fruit synchronously, conspicuous red or light yellowish fruits were used as targets for observation. Observations were made from a distance through binoculars. The fruiting condition of each forest stand was classified into one of four categories on the basis of the density of fruiting dipterocarp trees: none, no fruiting trees found; minor, up to 25% of large dipterocarps bearing fruits; intermediate, 25% to 50% of large dipterocarps bearing fruits; and major, 50% or more of large dipterocarps bearing fruits. Meteorological data source We examined long-term records of rainfall and minimum temperatures at the nearest meteorological station to the study site (Pasoh Dua, 2 56 N, E) from January 1981 to April Rainfall was not recorded during February April 1981 or January February 1982, and minimum temperature was not recorded during January April 1981, January February 1982, January February 1986, May September 1986, July 1987 May 1988, August 1988, July October 1992, or after April To determine the annual patterns of rainfall and low temperature on the Malay Peninsula, we also obtained data on monthly rainfall and monthly lowest temperatures recorded at 16 stations in Malaysia (Malaysian Meteorological Service; For monthly rainfall, at least 20 years records were used for each region, with the exception of Langkawi and Muadzam Shah (13 yr) and Kuala Krai and Kuala Terengganu (15 yr). For monthly minimum temperatures and lowest temperatures recorded, at least 15 years records were used for each region, with the exception of Langkawi and Muadzam Shah (13 yr). The El Niño Southern Oscillation (ENSO) is the major cause of interannual climate variation in the tropics. The Southern Oscillation Index (SOI), which is associated with El Niño (negative SOI) and La Niña (positive SOI), was provided by the Japan Meteorological Agency. Five-month moving averages of SOI were calculated to show the supra-annual climate pattern. We defined three ESNO phases: El Niño (SOI 0.5), normal ( 0.5 SOI 0.5), and La Niña (SOI 0.5). RESULTS Mass flowering and meteorology at the Pasoh Forest Reserve From 1980 to 2002, at least 11 flowerings were observed in the reserve (Fig. 1). Because the magnitude of flowering varied from less than 10 65%, we classified the flowerings into two categories, mass (major) and sporadic (minor) flowerings, on the basis of flowering densities of individual

3 July 2003] NUMATA ET AL. MASS FLOWERING ON THE MALAY PENINSULA 1027 Fig. 1. Climatic indicators and flowering events in the Pasoh Forest Reserve from 1980 to (a) Five-month average Southern Oscillation Index (SOI). The shaded line shows the normal ENSO phase ( 0.5 SOI 0.5). (b) Daily minimum temperature at Pasoh Dua, 5 km from the Pasoh Forest Reserve. Arrows indicate drops in minimum temperature to below 20 C before flowerings. (c) Monthly rainfall at Pasoh Dua. In this region, obvious flowerings were reported in 1976, 1981, 1983, 1985, 1986, 1987, 1989, 1990, 1996, 1999, 2001, and 2002 (Table 1). Shaded bars indicate mass flowering and hatched bars indicate sporadic flowering. trees and species. There was no regular periodicity in the occurrence of either mass or sporadic flowerings. In the years of mass flowering, we observed more than 40% of individual trees flowering, and sequential flowering by some Shorea species. Among all flowerings, flowering intervals varied from 1 to 6 yr, whereas in mass flowerings they varied from 4 to 7 yr. The timing of the flowerings had a clear seasonality (Fig. 1). Of the 11 flowerings, six occurred from February to April (1983, 1987, 1989, 1990, 1996, 2002); we regarded these as spring flowerings. The other five occurred from August to October (1981, 1985, 1986, 1999, 2001); we regarded these as autumn flowerings. We found plausible meteorological cues for mass flowering in drops in minimum temperature and low rainfall in massflowering years (Fig. 1). The correlation between monthly minimum temperature and rainfall was statistically significant (r 0.18, P 0.01, Pearson s correlation), but not strict. In all but the most recent mass flowering years (data not available for 2002), at least two periods of low minimum temperature ( 20 C) were found before flowerings (Fig. 2). In 1981, some weak drops in minimum temperature occurred. In 1985, there was one weak drop and one strong drop. In 1989 and 1996, two moderate drops occurred before flowering. The number of months during which low temperature ( 20 C) was recorded (59) was more than the number of flowerings (Fig. 1). Low temperatures were recorded in 41 mo in the 1980s, 16 mo in the 1990s, and 2 mo in the 2000s. Although the flowerings occurred during any ENSO phases, the drops of minimum temperature in flowering years mostly occurred in the normal phase (Fig. 1). In addition, both positive and negative changes in SOI occurred during the putative periods of floral induction in all mass flowerings. In the five mass flowerings, four putative periods of floral induction (in 1981, 1985, 1996, and 2002) occurred in the normal phase and one (in 1989) in the La Niña phase. In the sporadic flowerings, most putative periods of floral induction occurred in the normal phase, except in 1983 (El Niño) and 1999 (La Niña). We compared rainfall and minimum temperatures between the normal, El Niño, and La Niña phases. Monthly rainfall and lowest temperature did not significantly differ among the ENSO phases (rainfall: F 2, , P 0.73, ANOVA; lowest minimum temperature: F 2, , P 0.08, ANO- VA), but we found that the mean minimum temperature during the normal phase (21.8 C) was significantly lower than during the El Niño (22.2 C) and La Niña (22.2 C) phases (F 2, , P 0.02, ANOVA with Scheffé s post hoc test). Figure 3 summarizes the mean annual patterns of minimum temperature, rainfall, and flowering in the Pasoh Forest Re-

4 1028 AMERICAN JOURNAL OF BOTANY [Vol. 90 Fig. 3. Annual pattern of rainfall (bars), minimum temperature (line), and flowering types in the Pasoh Forest Reserve. In the autumn type, low temperature and drought occur in June, July, or August and flowering starts from August or September. The southwest monsoon (S.W.M.) from Sumatra or the Indian Ocean prevails from May to September. In the spring type, low temperature and drought occur in December, January, or February, and flowering starts from February or. The northeast monsoon (N.E.M.) from East Asia prevails from November to. Fig. 2. Daily minimum temperature (circles), daily rainfall (bars), and mass flowerings in the Pasoh Forest Reserve in 1981, 1985, 1989, and Arrows indicate drops in minimum temperature to below 20 C. The striped bars below the graphs indicate drought periods during the putative period of floral induction. serve. The bimodal patterns of rainfall and minimum temperature are explained by the northeast and southwest monsoons. One putative period of floral induction occurred during the first minimum of annual rainfall and minimum temperature, caused by the northeast monsoon in December February. The other putative period occurred during the second minimum of annual rainfall and minimum temperature, caused by the southwest monsoon in May July. Climate and mass flowering on the Malay Peninsula Although temperatures on the Malay Peninsula are generally uniform throughout the year, the peninsula can be classified geographically into two regions by its seasonal patterns of low-temperature records ( 20 C) (Fig. 4). Throughout the peninsula, there is a distinct pattern of low temperatures between December and February. Further, at 10 meteorological stations in the central and southern parts of the peninsula (closed squares in Fig. 4), the data show another season of low temperatures ( 20 C) in one or more of the months of June, July, and August. The rainfall pattern was also divided into two zones with monomodal and bimodal seasonal distributions (Fig. 4). The west and south have two distinct minima of monthly rainfall, in December February and June August. The northwest and northeast show one minimum, in December February. Although there were some exceptions, low temperatures ( 20 C) tended to occur during the minima of monthly rainfall (Fig. 4). We compared the geographical distributions of mass flowerings between spring 1996 and autumn 2001 (Fig. 5a, b). In 1996, mass flowering started in mid- and mature fruits fell in August and September. In 2001, mass flowering started in mid-august, and mature fruits fell in January In 1996, forests with large and intermediate-sized mass flowerings (87% of all forests) were observed throughout the peninsula (Fig. 5a), except in the central part of the west coast. In contrast, in 2001, large and intermediate-level mass flowerings were distributed only on the central west coast, on the central plain, and in the southeast (Fig. 5b). DISCUSSION Our results are consistent with two earlier hypotheses that focus on prolonged drought (Burgess, 1972; Medway, 1972; Appanah, 1985) and low temperature (Ashton et al., 1988; Yasuda et al., 1999). This consistency suggests that strong radiative cooling under cloudless conditions during drought is necessary for the occurrence of flowering. However, these conditions are not always followed by flowering events. Other factors such as resource status may influence flowering. Temporal pattern of mass flowering On the basis of herbarium specimens, Ashton et al. (1988) reported that flowering occurs in the second quarter on the Malay Peninsula and in

5 July 2003] NUMATA ET AL. MASS FLOWERING ON THE MALAY PENINSULA 1029 Fig. 4. Geographical patterns of monthly lowest temperature recorded (lines) and monthly rainfall (bars) at 16 meteorological stations on the Malay Peninsula. Regions are categorized by monomodal (open squares) or bimodal (solid squares) patterns of abnormal low temperatures ( 20 C). Scales on each graph indicate monthly rainfall (in millimeters) on the right axis and temperature (in degrees Celsius) on the left axis. Months in which abnormal low temperature occur are emphasized in boxes. eastern Borneo, but in the fourth quarter in south Sumatra and western Borneo. In contrast, we found that flowering occurs in both the second and fourth quarters on the Malay Peninsula. We attribute the seasonality of mass flowering to the annual pattern of rainfall and minimum temperature in this region (Fig. 3). This annual pattern is explained by the northeast monsoon from Asia and the southeast monsoon from the Indian Ocean (Dale, 1959). Thus, the monsoons are the key to understanding the seasonal and geographical patterns of abnormal low temperature.

6 1030 AMERICAN JOURNAL OF BOTANY [Vol. 90 Fig. 5. Examples of flowering status of forest stands in peninsular Malaysia in June July 1996 (a) and November 2001 (b). Each forest site was scored for the presence of young fruit. Data on fruiting status in June July 1996 (a) were derived from Yasuda et al. (1999) and revised for this study. In earlier studies of mass flowering and ENSO on the Malay Peninsula, in Borneo, and in Sumatra, no simple association was found (Burgess, 1972; Ashton et al., 1988; Curran et al., 1999; Yasuda et al., 1999; Wich and van Schaik, 2000; Sakai, 2002). Strong relationship between mass flowering and ENSO periods was reported in Borneo and eastern Malay Peninsula (Ashton et al., 1988; Curran et al., 1999; Wich and van Schaik, 2000), but most of the earlier studies analyzed mass flowering and ENSO based on the yearly data without considering the time lags from floral trigger to fruit dispersal (Yasuda et al., 1999). We found that mass flowering was triggered during the absence of El Niño episodes in the focal forest. This is explained by the fact that El Niño episodes generally decrease rainfall but increase average temperatures on the Malay Peninsula (Yasuda et al., 1999). However, there is not enough evidence to explain how the ENSO causes climate abnormalities in this region. Studies of the mechanism of ENSO and its effect on regional meteorology will be important to clarify its association with mass flowering. Another possible factor, the dipole mode in the Indian Ocean, should also be investigated (Saji et al., 1999). Spatial and temporal patterns of mass flowering on the Malay Peninsula The Malay Peninsula generally has a distinct minimum of monthly rainfall in December February (Dale, 1959). Further, as at the study site, the central and southern parts of the peninsula have another minimum in June August, owing to the southwest monsoon. In the northern part of the peninsula, a westerly upper wind generally blows (at 850 hpa) in June and July, but in the southern part a southwesterly upper wind blows, owing to the location of Sumatra (Matsumoto, 1992). Differences in the upper wind fields in June July probably account for the differences in rainfall and low temperature between the northern and other areas. Thus, we can hypothesize that there are two seasonally and geographically different types of mass flowering on the Malay Peninsula. This hypothesis anticipates two potential seasons when mass flowering is likely to occur in the southern part of the peninsula. The geographical distribution of mass flowerings in 1996 and 2001 strongly supports our hypothesis. A continuous examination of flowering distribution would clarify the spatial pattern of mass flowering. There is some evidence of two obvious seasons for mass flowering in the central and southwestern regions. Throughout the Malay Peninsula one distinct peak of flowering in the second quarter of the year (spring) was observed, and in several southern regions two mass flowering seasons were observed (Burgess, 1972; Wycherley, 1973; Yap and Chan, 1990). These classifications may provide useful information for the prediction of mass flowering on the Peninsula. Conclusion Our hypothesis provides a theory on spatial and temporal patterns of mass flowering on the Malay Peninsula. This predicts that strong drought seasons in spring and autumn are essential for mass flowering on the Malay Peninsula during the absence of El Niño episodes. Furthermore, our results imply that regeneration of the dominant canopy trees

7 July 2003] NUMATA ET AL. MASS FLOWERING ON THE MALAY PENINSULA 1031 in Southeast Asian forests rely on a delicate balance between the dry winter monsoon from the northeast and the dry summer monsoon from the southwest. Recently, effects of global warming are among the most pressing problems facing various ecosystems. Even though small increments in temperature are predicted to have relatively little influence on vegetation change in tropical regions, rapid global climate change might result in changes in plant phenology and could potentially have serious consequences for plants as well as animals that depend on periodically available plant resources (Corlett and LaFrankie, 1998). Pollinators and seed predators may be especially vulnerable to changes in plant resources that will accompany changes in mass flowering. Therefore, collaboration of scientists in fields such as reproductive biology, plant physiology, entomology, zoology, and meteorology is needed to confront the problem (Sakai, 2002). We strongly underscore the importance of long-term surveys of flowering and climate for the prediction of future changes in the tropical rain forest ecosystems of Southeast Asia. LITERATURE CITED APPANAH, S General flowering in the climax rain forests of South- East Asia. Journal of Tropical Ecology 1: ASHTON, P. S Dipterocarpaceae. Flora Malaysiana Series 1. Spermatophyta 9: ASHTON, P. S., T. J. GIVINISH, AND S. APPANAH Staggered flowering in the Dipterocarpaceae: new insights into floral induction and the evolution of mast fruiting in the aseasonal tropics. American Naturalist 132: BURGESS, P. F Studies on the regeneration of the hill forests of the Malay Peninsula. Malaysian Forester 35: CORLETT, R. T., AND J. V. LAFRANKIE Potential impacts of climate change on tropical Asian forests through an influence on phenology. Climatic Change 39: CURRAN, L. M., AND M. LEIGHTON Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting Dipterocarpaceae. Ecological Monographs 70: CURRAN, L. M., I. CANIAGO, G. D. PAOLI, D. ASTIANTI, M. KUSNETI, M. LEIGHTON, C. E. NIRARITA, AND H. HAERUMAN Impact of El Niño and logging on canopy tree recruitment in Borneo. Science 286: CURRAN, L. M., AND C. O. WEBB Experimental tests of the spatiotemporal scale of seed predation in mast-fruiting Dipterocarpaceae. Ecological Monographs 70: DALE, W. L The rainfall of Malaya. Journal of Tropical Geography 13: ISAGI, Y., K. SUGIMURA, A. SUMIDA, AND H. ITO How does masting happen and synchronize? Journal of Theoretical Biology 187: JANZEN, D. H Tropical blackwater rivers, animals, and fruiting by the Dipterocarpaceae. Biotropica 6: LAFRANKIE, J. V., JR., AND H. T. CHAN Confirmation of sequential flowering in Shorea (Dipterocarpaceae). Biotropica 23: MANOKARAN, N., AND M. D. SWAINE Population dynamics of trees in dipterocarp forests of peninsular Malaysia. Forest Research Institute Malaysia, Kuala Lumpur, Malaysia. MATSUMOTO, J The seasonal changes in Asian and Australian monsoon regions. Journal of the Meteorological Society of Japan 70: MEDWAY, F. L. S. L Phenology of a tropical rainforest in Malaya. Biological Journal of the Linnean Society 4: NORTON, D., AND D. KELLY Mast seeding over 33 years by Dacrydium cupressinum Lamb. (rimu) (Podocarpaceae) in New Zealand: the importance of economies of scale. Functional Ecology 2: NG, F. S. P Gregarious flowering of dipterocarps in Kepong, Malayan Forester 40: NUMATA, S., N. KACHI, T. OKUDA, AND N. MANOKARAN Chemical defences of fruits and mast-fruiting of dipterocarps. Journal of Tropical Ecology 15: SAJI, N. H., B. N. GOSWAMI, P.N.VINAYACHANDRAN, AND T. YAMAGATA A dipole mode in the tropical Indian Ocean. Nature 401: SAKAI, S General flowering in lowland mixed dipterocarp forests of South-East Asia. Biological Journal of the Linnean Society 75: SAKAI, S., K. MOMOSE, T. YUMOTO, T. NAGAMITSU, H. NAGAMASU, A. A. HAMID, AND T. NAKASHIZUKA Plant reproductive phenology over four years including an episode of general flowering in a lowland dipterocarp forest, Sarawak, Malaysia. American Journal of Botany 86: SYMINGTON, C. F Forester s manual of dipterocarps. Penerbit Universiti Malaya, Kuala Lumpur, Malaysia. THOMAS, S. C Inter-specific allometry in Malaysian rain forest trees. Ph.D. dissertation, Harvard University, Massachusetts, USA. TOY, R. J., AND A. G. MARSHALL Fruiting phenology and the survival of insect fruit predators: a case study from the South-East Asian Dipterocarpaceae. Philosophical Transactions of the Royal Society of London B 335: WHITMORE, T. C Tropical rain forest of the Far East. Oxford University Press, New York, New York, USA. WICH, S. A., AND C. P. VAN SCHAIK The impact of El Niño on mastfruiting in Sumatra and elsewhere in Malesia. Journal of Tropical Ecology 16: WYCHERLEY, P. R The phenology of plants in the humid tropics. Micronesia 9: YAP, S. K., AND H. T. CHAN Phenological behaviour of some Shorea species in peninsular Malaysia. In K. S. Bawa and M. Hadley [eds.], Reproductive ecology of tropical forest plants. United Nations Educational, Scientific and Cultural Organization, Paris, France. YASUDA, M., J. MATSUMOTO, N. OSADA, S. E. ICHIKAWA, N. KACHI, M. TANI, T. OKUDA, A. FURUKAWA, A. R. NIK, AND N. MANOKARAN The mechanism of general flowering in Dipterocarpaceae in the Malay Peninsula. Journal of Tropical Ecology 15: