Biotic Interactions in Tropical Rain Forest - Update on progress (March 2002)
The onset in September of a mast flowering event at Sepilok has resulted in a flurry of research activity on the reproductive ecology of the dipterocarps. It is just such an event that we had been preparing for for the past couple of years and as soon as flowers were noticed Richard Thewlis began placing lines in flowering trees to gain access to the canopy in order to sample flowers for studies on pollination success. During a mast flowering event many species of trees flower more or less synchronously over a period of about six to eight weeks, having remained sterile for several years. Given the very narrow window of opportunity it was necessary to act quickly and Richard's training and prior practice of rope climbing to gain canopy access proved essential. Richard, together with a local tree climber and supported by the postdoc Colin Maycock, was able to locate and collect flowers from the canopies of 2-6 trees each day and occasionally more. The massive size of trees necessitated ascending into canopies exceeding 60m in height and then edging along the main branches to reach the flowers that were presented at the outermost ends of the smallest branches.
Flowers were collected principally from species of the genus Shorea as it is these that most commonly flowered. Sample sizes of trees varied considerably reflecting the wide variation in population size and abundance of flowering individuals (only a fraction of all individuals actually flowered). While we had originally intended to pursue an assessment of pollination success of dipterocarps from other genera this proved to be impossible given the time and logistical constraints. Particular problems were presented by Parashorea tomentella, a huge and expansive tree which, owing to its architecture was difficult, and indeed dangerous, to climb anywhere beyond the central trunk. Species of the genus Dipterocarpus also proved difficult in that at any one time only a few large flowers are produced and securing sufficient sample sizes proved difficult and exceedingly time consuming. Ultimately we were able to collect flower samples from over 160 trees representing eleven species of Shorea and Hopea.
The flowers were collected using a pruning pole and transferred directly into Formalin: Acetic Acid: Alcohol (FAA) solution which fixes the plant tissue. Only flowers that had been open and exposed to pollinators for at least 12 hours were collected (flowers of these species open at dusk and rarely retain their corollas for more than 24 hours). Pollination success will be determined using fluorescent microscopy in the laboratory at Silwood Park over the coming months. The advantage of this technique is that it allows detection of the pollen tubes indicating the successful transfer of compatible pollen grains to the stigmas of the flowers. Inbreeding among these trees is limited by the presence of a self-incompatibility mechanism that acts to prevent germination of self-pollen. Thus the presence of pollen tubes is a good indication that pollen has been received from a neighbouring conspecific. The results emerging from the analysis of pollination success will be correlated with density of adult trees and nearest neighbour distances to flowering conspecifics which will test the hypothesis that fruit set is a negative function of tree isolation based on the assumption that pollinators move between isolated trees less frequently than near neighbours.
Building upon the flowering work we intended to follow patterns of fruit development, abortion and predation for all individuals from which flowers were sampled, as well as for other trees from which we were unable to sample flowers for practical or time constraints. To sample developing fruit four 'seed traps' were placed under the canopy of each study tree as soon as flowering was detected. While labelled as 'seed traps' these 1m2 traps were actually used to sample and quantify the production of flowers, aborted fruit, and mature fruit throughout the reproductive period. Material was collected from each of these traps every three to four days, air dried in an air-conditioned office and sorted into buds, flowers, young aborted fruit, larger aborted fruit and mature fruit. The number of buds, flowers, and fruit from each of these samples is being counted to be able to relate flower production with fruiting success, as well as to quantify the impacts of insect predation and to relate the pollination studies (above) to ultimate seed set. Mature fruit is still falling at the time of writing although the process of collection from seed traps is nearing its conclusion. The process of sorting and quantifying the samples is well underway and despite the slow start, owing to the immense numbers of flowers obtained (several thousand from a 4m2 area), this part of the analysis is speeding up as samples no longer include flowers and consist mainly of some fruit.
The remarkable results that are being obtained is the immense apparent 'wastage' of flowers compared to the number of mature fruit that are ultimately formed. Typically less than 30 fruit are being obtained from over ten thousand flowers, with many flowers initiating further development only to be aborted at an early stage. This could be related to poor pollination success which should become apparent with further laboratory analysis.
Developing and mature fruit that have been predated by insects are readily distinguished by the leakage of resin from the oviposition hole. These fruit have been placed in small simple chambers to allow development of any insect larvae to adults to facilitate their identification. To date most predation appears to have been caused by micromoths (Tortricidae and Pyralidae) which we hope to identify further with assistance from taxonomic expertise in the UK. Less frequent is damage due to weevils (probably Alcidodes spp. and Nanophyes spp.) which is surprising given that earlier studies in Sabah and peninsular Malaysia in similar forest types indicated that weevils were by far the most important pest of dipterocarp fruit. Insects are continuing to emerge from fruit and our preliminary findings may indicate no more than that weevils have a longer period of development compared to moths.
Mature fruit that has apparently escaped predation by insects appears to suffer from mammal predation, both pre-dispersal predation by monkeys and postdispersal predation probably by rodents and, to a lesser extent, pigs. Predation even extends to the few remaining fruit that do succeed in producing cotyledons and the first true leaves. It is becoming apparent, through work in collaboration with Colin Maycock, that the combined effects of insect and mammal predation on fruit and very young seedlings is causing almost 100% mortality of progeny. Such massive impact of predation does not appear to be typical of other masting events where sufficient number of fruit and seedlings have survived to produce a scattering of seedlings (now mostly between 30 and 60cm tall) in the vicinity of the study trees.
Our conclusion regarding the current masting event is that it is relatively minor in extent compared with previous events, an observation that is supported by the fact that less than 40% of individuals of most species that were part of the masting event actually flowered. This raises a number of questions regarding the timing of masting events and the cues that control their occurrence, as well as the ecological implications to individual trees of flowering in a relatively poor masting event.
Work over the coming months will be concentrated on completing the fieldwork (due to finish by the end of March) and beginning to assess pollination success through pollen tube analyses unfer fluroesecnec microscopy. A full analysis of the data obtained will be undertaken soon after completion of the pollination work.
Jaboury Ghazoul, Imperial College London
| 
AREA OF INTEREST:
Home
Introduction to the project
