Vegetatio 121: 189-191, 1995. T. Hirose and B.H. Walker (eds). Global change and terrestrial ecosystems in monsoon Asia (~)1995 Kluwer Academic Publishers. Printed in Belgium.

189

Abstracts

THE C H A R A C T E R I S T I C S OF WARM T E M P E R A T E FORESTS IN M O N S O O N ASIA A N D ITS R O L E IN THE G L O B A L C H A N G E Zongwei Feng & Xiaoke Wang Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100080, China The warm temperate forest in Monsoon Asia is located between 25 and 40 ° latitude and mainly distributed in China. A little in Korea and Japan. First of all, it must be mentioned that there are some geographical characteristics that affect the distribution, structure and function of the forest: (1) the range lie in the eastern part of Eurasia continent and neighbors with Pacific ocean that are the largest of ones of the world; (2) the climate is controlled by Siberia High Pressure in winter and by Pacific Low Pressure in summer, so that weather changes obviously with seasons; (3) there exists topographical and climatic gradient from coast to inner of the continent. The warm temperate forest in Monsoon Asia consists of deciduous forest, mixed deciduous and evergreen broad-leaved forest from north to south. The latter two types are considered as subtropical forest by some Chinese botanists. The forest is limited not less than 100 o longitude because of aridity and height in the west China. However, the forest distribute scatteredly and most of it are secondary forest because the land of this area has been cultivated intensively in order to sustain more than 800 million people, about more than 15% of that in the world. The dominant species are Quercus, Betula, Acer and Populus in deciduous forest and Cyclobalanopsis, Castanopsis and Lithocarpus in mixed and evergreen broad-leaved forest. It is necessary to note that there is a large area of artificial forest constructed since 1949 that consist of Populus, Pinus massoniana and Cun-

ninghamia lanceolata. The biomass and productivity of warm temperate forest vary largely from island oceanic weather to inner continental weather. Some author believed that warm temperate forest is a sink of CO2 in the atmosphere,

But as the preliminary results of our studies, the warm temperate forest in Monsoon Asia has a limited, steady area. Most of the forest locates in mountainous range and has been mature, so the forest has no important role in influencing the budget of CO2 in the atmosphere. Here it is necessary to mention that the agroforestry ecosystems, called eco-engineering project by Chinese scientist, have been developing rapidly that will increase carbon pool and absorb CO2 from the atmosphere.

T R O P I C A L RAIN FORESTS OF THE INDO-MALAYAN REGION N. Manokaran Forest Research Institute Malaysia, Kepong, Selangor 52109, Kuala Lumpur, Malaysia

Keywords: Biological diversity, Conservation, Human impacts, Indo-Malayan region, Tropical rain forests The tropical rain forests of the Indo-Malayan region stretch from Sumatra in the west, through the Malay Archipelago, to New Guinea in the east. They are evergreen, hygrophilous in character, at least thirty metres high, rich in thick-stemmed lianas, and in woody as well as herbaceous epiphytes. In this paper, the rain forest types in the IndoMalayan region are described, and the dominance of the family Dipterocarpaceae discussed. Stratification in the dominant forest type, the dipterocarp forest, is also described and an example given of the woody composition of this stratification. The high biological diversity and the high endemicity in these rain forests are illustrated with examples. Information on recruitment and mortality patterns in tree species in these forests is also provided. Finally the human impacts on tropical rain forests in this region are traced with reference to agriculture, including shifting cultivation, and to logging for timber, and these issues discussed in the overall context

190 of conservation and protection of this ancient ecosystem.

and behavior of caterpillars of a generalist and a specialist will be compared.

GROWTH, REPRODUCTION AND DEFENSE

S C A L I N G UP F R O M T H E PATCH TO T H E LANDSCAPE L E V E L

Fakhri A. Bazzaz

Ian R. Noble

Harvard University, Department of Organismic & Evolutionary Biology 16 Divinity Avenue, Cambridge, MA 02138, USA

Ecosystem Dynamics Group, Research School of Biological Sciences Australian National University, Canberra ACT2601, Australia

The role of forests in the carbon cycle and their ability to respond to global change (temperature and CO2 rise and increased N-deposition) is a critical question for science and policy. We have examined the patterns of net ecosystem exchange of CO2 between a deciduous forest and the atmosphere and have estimated CO2 flux from the soil. We have also examined the growth and allocation in representative species of the forest under combinations of elevated CO2 nutrient and light levels by simulating present and future gap and understory environments. Growth was initially enhanced in all species, but there were differences among them. Also, the responsiveness to elevated CO2 in these species declined differentially after prolonged exposure to elevated CO2 and in some species become negligible by the third growing season. We also found that the nitrogen/lignin ratio in live leaves and leaf litter declines when plants are grown under elevated CO2 conditions which lead to decrease decomposition rates. Together with reduction in responsiveness after prolonged exposure, this negative feedback can result in only small changes in ecosystem productivity and carbon sequestration by forests in the future. Allocation to reproduction is sensitive to CO2 levels, but the available results are limited and somewhat inconsistent. Using the annualAbutilon, we developed a population simulation model, based on experimental data, to test effects of elevated CO2 seed production, seed bank longevity, the percentage of seed germination, and the initial density on the fate of populations. We found that elevated CO2 leads to chaotic behavior and extinction, while ambient CO2 leads to stable populations after a few generations. The change in tissue chemistry, especially the reduction in nitrogen concentration and the increase in carbon-based defenses, reduce the ability of insect herbivores to grow and complete their life cycles. Growth

A landscape is an interacting set of contiguous patches covering an area of a few km 2 to thousands of km 2. Many critical processes that dominate the interactions between the biotic, abiotic and human processes emerge only at this scale (e.g. fire spread, nutrient redistribution, meta-population, human settlements). It is at this level that most decisions about environmental management are made (e.g. farming, coupe selection, urban planning). Few models exist of the important processes at this scale. The reductionist tendency in the natural sciences has led to most models being designed to deal with much smaller scales, e.g. hypothetical points or small patches of vegetation. A 'patch' is an idealized location in which it can be assumed that many individual plants exist and a wide range of interactions such as shading, competition for water and nutrients etc. occur, but which is small enough that every plant fully interacts with every other plant. Thus, detailed spatial relationships between the individual plants can be ignored. Some modellers have tackled larger spatial units by linking together a series of patch models with a selected range of inter-patch interactions. The heavy computational loads of this approach means that the processes and interactions that emerge at landscape scales via such models have been little explored. Another approach is to define simple rules about the interactions between the elements of the landscape and to explore the impact of changing the intensity or frequencies of some of the interactions. This approach is limited by our ability to define the rules. Experimental studies at the landscape scale are difficult and observational studies are usually confounded by numerous uncontrolled variables. Another approach has been to accept the need to synthesize landscape models from models derived from smaller scales. This approach requires a suite of

191 models of vegetation change of different resolution. The models are linked in the sense that the output from a more complex (high resolution) model can be used to parameterize a simpler model. For example, we may explore the biological effects of different fire regimes in a detailed patch model and from this estimate the probabilities of mortality of different species. These life-history data can be used to parameterize a simpler model that can be applied to a gridded land-

scape. We have also explored transforming moderately complex models into a more mathematically tractable semi-Markov processes. This allows us to use search and optimisation techniques such as dynamic programming to explore the best tactics to achieve particular management goals. The semi-Markov representation of the model reduces running of the model to accessing a lock-up-table. This means that they may be incorporated into a GIS as another data layer.