No.1

Managing coppice woodlands in traditional rural landscapes:

an international perspective

R. Lafortezza¹, G. Sanesi¹, O. Ciancio², S. Nocentini², M. Yokohari³ and M. Amati¹

1.Dept. of Plant Production Science, University of Bari, Italy

2.Dept. of Science and Forestry Technologies, University of Florence, Italy

3.Graduate School of Systems and Information Engineering, University of Tsukuba,

Key Words: landscape ecological management, coppice woodlands, Satoyama landscaoe, Mediterranean landscape

Traditional rural landscapes generally revolve around coppice woodlands. In longer-established agricultural countries, like Japan and Italy, the practice of coppicing trees has created unique ecosystems that often have a potentially high conservation value. Many coppice woodlands that remain in these countries have become neglected and require management in order to retain their character and biodiversity value. On the basis of this premise, in this paper we propose a comparative analysis of woodlands in selected areas of Japan and Italy. Traditional Satoyama landscapes are compared and contrasted with Mediterranean coppices in terms of average patch size and shape, species diversity, proximity to highly dense urban areas, etc. For these landscapes, we propose a landscape ecological approach to woodlands management which includes principles and guidelines for preserving their multi-functionality and pattern.

No.2

Landscape analysis at national scale using remote sensing data

- A case study in Cambodia -

Hideki Saito

Kyusyu research center, FFPRI

Key Words: Land cover, remote sensing, forest management

This study aims to produce a land cover map from remote sensing data and evaluate characteristics of forest distribution using landscape metrics. The study area was selected whole Cambodia. In this country, the sustainable forest management has been required for adequate use of forest resources, natural forest conservation and wildlife protection. Moreover the forest plays an important role in water cycle of Mekong River and Tonle Sap Lake. Water cycle management is a serious matter for this country, because rice is one of the main crops and an inundation was frequent occurred. SPOT VEGETATION (VGT) 10-days composite data (S10 product) was used for land cover mapping. SPOT VEGETATION observed 2,250 km wide of ground surface with 1 km spatial resolution. Observation period is more than once a day. Landsat-7 data, which has a high spatial resolution sensor, was used as reference data. Mapping accuracy was evaluated using existing land cover map produced by aerial photographs. The Normalized Difference Vegetation Index (NDVI), which is related to the proportion of photosynthetically absorbed radiation, was calculated from reflectances from the visible and near infrared SPOT VEGETATION channels. Then cloud and haze were reduced by time-series data analysis. Classification was performed based on seasonal NDVI change. Evergreen forest, deciduous forest, shrub and grass, urban and water were distinguished on the classification image. Forest distribution properties were evaluated by calculating several landscape metrics based on the land cover map. These landscape metrics could be directly compared between different areas in Cambodia, since the land cover was investigated by one dataset. Therefore these metrics are useful for policy maker to decide landscape management plan. This approach could be applicable to MODIS, which has a 250 m spatial resolution sensor for more accurate assessment and NOAA-AVHRR, which has long-term data set for retrieving long-term landscape change history.

No.3

Disturbance regimes and silvicultural methods in mountain forests of

southern Siberia fragmented by clear-cuts and wildfires

Igor M. Danilin and Sergei K. Farber

V.N. Sukachev Institute of Forest, Russian Academy of Sciences, Siberian Branch, Akademgorodok, 50/28, Krasnoyarsk, 660036, Russia

Key Words:

Mountain forests in Southern Siberia grow at or near southern boundary of the boreal zone in Central Asia. Sub-boreal, subarid and strongly continental climates are the major limiting factors for the forest coenoses in this area. Here, mountain forest ecosystems develop on seasonal frost soils and are surrounded by vast steppe and forest-steppe areas and uplands.

Developing silvicultural methods and criteria for stability of mountain forest ecosystems is necessary for appropriate forest resource management. It is therefore important to obtain complex biometric characteristics of forest stands of different ages and thoroughly study their structure and biological productivity. This information would be also vital for understanding the geographic variations in forest succession processes in disturbed mountain forest ecosystems.

Complex estimation of forest ecosystems¡¯ dynamics based on detailing local peculiarities of structure and growth of stands and disclosing general geographical and biological regularities of preserving their biodiversity and sustaining stability are discussed on example of mountain forests in Tyva Republic and southern regions of Krasnoyarsk territory.

Dynamics of organic matter and biodiversity serve as indices for stability of forest ecosystem. Examination of production dynamics related to stand structure and growth makes it possible to define permissible limits for affecting ecosystem and the extent and nature of their effect on the environment.

Ecosystem here is understood as open self regulated aggregate of interrelated natural formations, each of one has a beginning, proper existence and the end. After breakdown of ecosystem its place is occupied by analogous or quality different ecosystem. If enough knowledge exist for the state of affairs, process of the ecosystem's change can be predicted.

Disturbances and fragmentation of mountain forest ecosystems is a result of external influences. A possibility of beginning of disturbances is a result of ecosystem's openness. Ecosystems constantly experience different influences and this is normal for their existence. Negative consequences are neutralised by ability of self-regulation and restoration.

Stability of ecosystem is not infinite. Increasing of a force of influence leads to destruction and ecosystem completely or partially loses its territory. Earth surface is re-distributed in mountains. A part of lost territory resulted by external influence, where irreversible changes have happened and also dynamics of biological productivity could serve as a measure of ecosystem's stability.

Disturbances of mountain forest ecosystems can be classified as weak, moderate, medium, strong and extreme. This classification corresponds to any level of generalisation of ecosystem, beginning with a forest management compartment and ending by planet Earth, which is dynamic ecosystem successively including all others.

Comparing lost territory with force of influence and productivity parameters, conditions and limits of ecosystem's existence can be revealed and adequate and effective prognoses for their further evolution can be made.

No.4

Effect of different ecological restoration models on soil properties

in subtropical region of Southern China

Anand N. Singh* ^{1, 2}, Mingguang Li³, Dehui Zeng¹ and Fusheng Chen¹

1.Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China

2.Department of Botany, Banaras Hindu University, Varanasi 221 005, India

3.Institute of Botany, School of Life Sciences, Zhongshan University, Guangzhou, 510275,

Key Words: Rehabilitation, Heshan, Microbial biomass, Acacia mangium, Acacia auriculaeformis,

Eucalyptus exserta, Pinus massoniana, Soil development

This study was conducted at Heshan Hilly Land Integrated Experimental Station (112¡ã54'E, 22¡ã41'N) located in Heshan City in the central part of Guangdong Province, P.R. China. The experimental area is typical of low hills and small catchments. The soil textural class across the selected sites was oxisol developed from sand stone, with an acidic pH range 4.5-5.5. The climate of the region is characterized by subtropical monsoon and marked seasonal variation. We selected four monocultured plantations (Acacia mangium, Acacia auriculaeformis, Eucalyptus exserta and Pinus massoniana) and four mixed species plantations (legume with non-legume, deep rooted with shallow rooted, fast growing with slow growing species and evergreen with deciduous) for this study. As for comparison, we also selected a bare land, grassland and a natural secondary forest (NSF). The objective of the present study was to compare the effect of different ecological models on restoration of soil biological fertility across two contrasting growing period (dry and wet season). Therefore, the effect of planted species which improve soil fertility as indicated by physical characters viz., water holding capacity (WHC), texture, and bulk density (BD), and chemical characteristics such as pH, available nutrient, levels of mineral N and PO4--P, and rates of N-mineralization and nitrification was monitored. Development of soil microbial biomass and levels of microbial nutrients were also determined. In this study, the difference of soil physical characters especially texture, WHC were not statistically significant across selected ecological models (sites), although values were substantially different among all research sites. However, chemical characters showed a remarkable trend and reflected significantly different due to sites and seasons. Across sites, highest values of mineral-N, available-Pi, N-mineralization rate and microbial biomass C, N and P were found in NSF while lowest in bare land followed by grassland site, respectively. Results indicated that development of soil biological fertility was strongly integrated with growth and development of trees, species-specific nature and composition under different ecological restoration models in contrasting growing period. More specifically, leguminous species had resulted greater carbon, total nutrients (nitrogen and phosphorous) and levels of microbial biomass in single species plantation as compared with other species, however, corresponding values were substantially higher in the mixed plantations. Therefore, we concluded that mixed culture ecological restoration models might be used to facilitate soil restoration process on degraded land. Although it was short time observation, therefore, we suggest that long-term effects on soil properties should be considered as one of the criteria when selecting species for restructuring a forestland on degraded land.

No.5

Growth performance of some commonly grown agroforestry tree species

in an upland ecosystem:

Assessment of tree-site suitability using field survey,GIS and multivariate analysis

Robert G. Visco (1a), Bobby A. Crisostomo (2), Damasa M. Macandog (1b)

and Marc M. Delgado (1c)

(1a) Assistant Professor, Institute of Renewable Natural Resources; (1b) Assistant Professor, 1cResearch Assistant, Institute of Biological Sciences, University of the Philippines Los Ba¬Ðs, College, Laguna, Philippines

(2) Chief, Database Management, National Mapping and Resource Information Authority, Makati, Philippines

Key Words: tree suitability analysis; site assessment; GIS; multivariate analysis

Agroforestry is a sustainable alternative to "slash and burn" agriculture to address the increasing fragmentation of forest areas in upland ecosystems. Selecting potential agroforestry tree species and matching these with prevailing site conditions require a range of information to be evaluated. In this study, an attempt was made to assess the suitability of growing timber species (Gmelina arborea, Acacia mangium, Eucalyptus deglupta and Swietenia macrophylla) in Claveria, Misamis Oriental, Philippines, using information from field surveys, GIS and multivariate analysis.

A series of field surveys was conducted in the study site to gather tree biometric information, farm parcel characteristics, and soil fertility data. A parcel-tree growth database was developed to evaluate tree performance based on the actual growth of the trees as compared with the tree growth potential reported in literature.

Climatic and soil attributes of the farm parcels (air temperature, cloudiness, annual rainfall, relative humidity, elevation, aspect, slope, and soil fertility) were collected and mapped in GIS. The parcel-tree growth database was linked to GIS and analyzed using multiple linear regression to identify the significant indicators of tree growth for each of the tree species.

Results show that the most suitable areas for all the species in this study are in locations with aspects facing north to southeast and with flat to rolling (3-11ｰ) topography. S. macrophylla, G. arborea and E. deglupta grow well in temperatures ranging from 21-30ｰ C and A. mangium in elevations between 500 to 700 m asl. Other climatic and soil factors were identified to have varying relationships to tree growth.

The results of this study will be very useful in the selection of appropriate species for agroforestation within the study area.

No.6

Forest Cover Change Investigation in the Mount Pulag National Park, Philippines,

Using Remote Sensing and GIS

Angelito Turda¹, Masamu Aniya² and Melchor Deponio³

1.Graduate School of Life and Environmental Science, University of Tsukuba, Japan

2.Institute of Geoscience, University of Tsukuba, Japan

3.Department of Science and Technology, La Trinidad, Benguet, Philippines

Key Words: Mount Pulag National Park, Remote Sensing, GIS, Landsat TM, Supervised Classification, Deforestation

Mount Pulag National Park lies in the Grand Cordillera Central in the northern part of the Philippines at 16o 30¡¯ 36¡± N and 120o 50¡¯20¡± E covering an approximate area of 11,550 hectares. It is considered to be the last frontier of the Philippine mossy forest. During the past decade, a considerable amount of forest cover had been lost leading to some negative impacts such as loss of biodiversity and soil erosion. Mitigating the impacts entails the generation of relevant data and information on current land management and spatial-temporal land use changes.

The study aimed to quantify the spatial and temporal changes of forest cover in the Mount Pulag National Park from 1988 to 1998, and to identify the different variables such as road networks, drainage networks, residential houses, aspect, slope, and elevation associated with these changes. The study used several maps such as topographic maps, land use and forest type maps and two Landsat Thematic Mapper images taken on January 31, 1988 and January 10, 1998, respectively. The data were processed using image processing and Geographic Information System. Post-classification scheme was done based on the supervised maximum likelihood classification to map out forest cover changes. Thematic maps resulting from the single-time classification were compared with each other quantitatively. Land cover classes were lumped into forest and non- forest categories (binary format). Buffer distances were created and overlaid to the deforested area map to investigate the degree of relationship with deforestation by using distance from road, household, and river channel as indices. Landscape attributes such as slope, elevation, and aspect were also used to investigate the degree of relationship to deforested area. During the period of ten years, about 652 hectares were changed from forest to non- forest areas. Regression analysis was employed to determine the amount or degree of deforested area in relation to distance from the road, the distance from household, and the distance from river channel. Results showed that there was a high significant degree of relationship between distance from road, household, and river channel to the amount of area deforested. Most of the deforested areas were concentrated from 0- 40% slope, within 1700 - 2200 m above sea level and to the west to southern aspects.

No.7

Ecosystem approach for the landuse and forest management in tropics

Toshinori Okuda¹, Keiichiro Yoshida^1,3, Sinya Numata¹, Sen Nishimura¹,

Mariko Suzuki¹, Toshiaki Kondo¹ and Mazlan Hashim²

1 National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba 305-0053 Japan

2 Universiti Teknologi Malaysia, Skudai, Johor Bahru, 81310 Malaysia

3 Present affiliation: Yokohama National University

Key Words: integrated ecosystem assessment, rapid assessment, risk assessment, landscape zoning assessment, tropical rain forest

The development of ecosystem management on tropical forest ecosystem requires intensive studies of the ecological service values and goods of various types of landscape in tropics. Studies should begin with a review of the functional aspects of the forest, agricultural urban and watershed ecosystems in the target study area. Furthermore, new and more data should also be input to the unknown ecological services. In order to develop and implement the well prepared management plan, we definitely need database of ecological service values and goods which facilitate the analyses of the interactive relationship between the different types of ecological services. Studies of highly diverse biomes such that seen in tropics need tools that can optimize the values of the different types of ecological services, particularly conflicting ecological services, such as biodiversity and timber production or carbon sequestration. The ecological service database will allow the development of a risk assessment program, a landscape zoning plan, and other planning instruments that play important roles in ecosystem management.

No.8

The Ecological Consequences of Landscape Fragmentation:

From Islands to Continents.

Thomas R. Crow

USDA Forest Service, Washington, D.C.

Key Words: land use, landscape change, management, Japan, spatial pattern

Landscape fragmentation, the transformation of large habitats or land areas into smaller parcels, is commonly associated with human land uses. It has become an environmental and management issue with global implications. The major spatial processes associated with fragmentation generally increase the patch numbers, reduce the average patch size, reduce interior habitats, reduce the connectivity across the landscape, and increase the boundary length between patches in a landscape. These changes result in the loss of undisturbed or minimally disturbed habitats and the isolation and possible loss of species that depend on these habitats. It also results in the increase of disturbed habitats and an increase in the populations of organisms that are adapted to disturbance. The study of the spatial processes associated with fragmentation on islands such as Japan with its intense land use and long period of extensive landscape modification may provide useful insights for other areas in which the extent, intensity, and duration of fragmentation are less advanced

No.9

Linking Ecology to Landscape Hierarchies

Jiquan Chen, Sari Saunders, Kim Brosofske, and Thomas Crow

University of Toledo

Key Words: landscape, ecosystem, concept, processes, hypothesis, disturbance

Among many challenges facing ecologists within the arena of landscape ecology, the lack of sound theories and hypothesis-oriented investigations has been increasingly discussed as a persistent obstacle to advancing this scientific discipline. The central focus of our study was to emphasize the ¡°ecology¡± of landscapes rather than landscapes themselves. Using examples, we illustrate how an ecological focus can be achieved in conducting landscape-level research. Specifically, we have researched the importance of landscape elements such as roads, edges, streams, different types and ages of patches, etc. in shaping overall landscape functions such as production and biodiversity. We have also examined how ecology-landscape relationships may vary with scale. Concurrently, we have simulated various disturbances (natural and anthropogenic) and changes in climate to examine their effects on multiple ecological processes and functions. Our results are supported by a large database of field observations (e.g., vegetation sampling, microclimatic records, tower-based flux measurements) and experimental data primarily from a landscape in northern Wisconsin, USA. For example, we conclude that landscape-level carbon flux can be confidently predicted from its composition, if patch age, disturbance history, and ecosystem characteristics (e.g., foliar N, leaf area) are included in calculation of carbon fluxes. Finally, we propose a few scientific hypotheses for design and development of future research. Particular attention was given to developing alternative management scenarios by linking land mosaics to the ecological functions of landscapes.

No.10

History of woodland structure and composition

on the Inashiki Plateau of southern Ibaraki Prefecture, Japan

David S. Sprague and Nobusuke Iwasaki

National Institute for Agro-Environmental Sciences, Tsukuba, Ibaraki, Japan

Key Words: rapid survey maps, cartography, rural landscape, structure and composition of woodlands

The Rapid Survey Maps, the first maps surveyed by modern cartographic methods in Japan, provide detailed information about land use under traditional agriculture. The maps show that the rural landscape in southern Ibaraki Prefecture in the 1880¡¯s had included large blocks of woodland and grassland. The grasslands have now disappeared, and although some of these woodlands still remain, the structure and composition of woodlands continue to change today.

No.11

Biodiversity conservation and the Japanese forest management policy

Ken Sugimura

Forestry and Forest Products Research Institute, Research Information Center

Key Words: Biogeography, national park, pristine ecosystem, plantation, national forest, conservation

Japan has various forest vegetation types due to latitudinal and altitudinal extension, from subtropical to boreal and from the sea level to Alpine tundra. Also biogeographical diversity is great due to isolation from the continent and/or between island groups. There are a few jurisdictional systems that attempt to preserve some representative pristine forest ecosystem types over the country, such as national parks, wilderness areas and nature conservation areas. On the other hand, the department of forestry has encouraged extensive clear cutting of natural forests to replace them with plantations as well as deforestation for recreational use of forest areas. It was because they endeavored to maintain a self-supporting accounting system of the national forest. As a result, relatively untouched forest ecosystems cover only small area relative to its original size but for high mountain areas, so that they are fragmented to each other. In the meantime, the amount of domestic timber supply declined gradually since late 1960s to compose only about 20 percent of the total demand. Therefore, the government had to determine to give up the former system and make a fundamental change in the forest management plan, in which forests are divided into three types as the primary management objectives, (a) timber use, (b) land and water conservation, and (c) amenity and biodiversity. Yet, the change was not virtually fundamental, since the classification succeeded existing forest types, basically plantations, naturally regenerated secondary forests, etc. Reflecting these situations, forests on Amami Island, where many rare species are found, have been cut clear, before invasive mongooses expanded their distribution areas and contributed to further decline of the populations to be protected. Even though the government spends a large amount of budget for constructing forest roads and dams, they spare the minimum amount for biodiversity conservation despite of great public interests.

No.12

Advance and Perspective of Forest Landscape Ecology in China

Shirong Liu¹, Yong Lin² and Jianping Ge²

1.Chinese Academy of Forestry, Beijing 100091,P.R. China 100091

2.College of Life Sciences, Beijing Normal University, 100875 Beijing, P.R. China 100875

Key Words: China, landscape ecology, forest management, forest hydrological process

Historical development and current advances of forest landscape ecology in China are briefly reviewed, and followed by exploring future research direction and focus on application of forest landscape ecology to forest management, wildlife conservation and forest eco-hydrology. Extensive research work on structure, dynamics and planning of forest landscape in some typical forest regions in China are also covered. The main problems that forest landscape ecologists in China have faced include shortage of long-term historic data, difficulty of linking landscape pattern and landscape process, and lack of well-elaborated landscape model.

A case study of effects of forest landscape change on hydrological process in Minjing River, China is given as an example. The conception and structure of macro-scale eco-hydrological model, in particular, scaling effects in terms of definition and comparability of spatial scale and temporal scale is discussed. Modeling process integrated with geographic information system, development of high-resolution data sets including land use and vegetation cover, soil texture, potential evaportranspiration are introduced. Future research perspectives of forest landscape ecology are envisioned, including development of coupling model of forest landscape and ecosystem, field experimental approach establishing the relationship between landscape pattern and landscape process. Some challenging issues forest landscape ecology in terms of interaction of spatial pattern, ecological processes and function, methodology of scaling up, and coupling ecological process with hydrological process are discussed.

No.13

Improvement of the living space of wildlife through reforestation

in Thai Nguyen Province(Vietnam): a spatial analysis

Nguyen Van Sinh and Dao Thi Chau Ha

Institute of Ecology and Biological Resources, Vietnamese Academy of Science and Technology

Key Words: fragmentation, home range, wildlife, reforestation

Forest loss causes a direct reduction of the living space of the wildlife. Moreover, the resulted forest fragmentation makes also a lot of forest area unsuitable for the wildlife because of the small size of the fragmented forest patches. The increase of total edge length is the further negative consequence of the anthropogenic forest fragmentation, because many wildlife species do use only the core areas of the forest patches, so that an increase of the total edge length means also a decrease of the living space.

To see how the reforestation in the past contributed to the improvement of living space of wildlife, and to see its prospect in terms of future reforestation of the currently bared land, the spatial analysis was conducted for the following maps of Thai Nguyen province that is located in the North of Vietnam: the map of natural forest, the map of planted forest, the total forest map, the map of bare land, and the map of the total forest combined with bare land. The minimum home ranges of several wildlife species of the Thai Nguyen province were considered in this analysis.

No.14

Landscape disturbance models and the management of resource and fire

Chao Li, Hugh Barclay Jianwei Liu and Doug Campbell

Canadian Forest Service

Key Words: fire regime, SEM-LAND model, Saskatchewan, fire suppression simulation, fire management

Landscape disturbance models can be useful in assisting the management decision-making of forest resource and disturbances such as fire. This presentation will describe a case study on how the SEM-LAND model was used to simulate both natural and current fire regimes in central Saskatchewan. The model used the indexes and relationships summarized in the Canadian Forest Fire Weather Index System and the Canadian Forest Fire Behavior Prediction System to simulate the interactions among fire events, landscape structures, topography, and weather conditions. The impact of fire suppression operations on fire spread was simulated the following three stages: fire ignition, initial attack, and fire escape. Our results suggest that fire suppression could lead to longer fire cycles, smaller mean fire sizes, higher mean fire numbers per year, and increased mean forest age (time since fire). The relationship between mean forest age and annual area burned was estimated from the results of this model experiment. This relationship can be used to determine the ideal fire management target of annual area burned from expected future forest conditions. We then conclude that the SEM-LAND model was capable of simulating fire regimes in Saskatchewan, and the simulation results could assist forest managers in determining the annual allowable cut and the fire management target of annual area burned.

No.15

Responses of Bird Assemblages to the Abandoned Settlement Areas

in Thung Yai Naresuan Wildlife Sanctuary ; A Natural World Heritage,Thailand

Prateep Duengkae¹, Sompoch Maneerat², Anak Pattanavibool³ and

Dokrak Marod¹

1 Forest Biology Department, Faculty of Forestry, Kasetsart University, Bangkok 10900. Thailand.

2 Protected Area InnovationUnit,Department of National Parks,Wildlife and Plant Conservation, Bangkok 10900. Thailand

3 Biology Department,Faculty of Sciene, Mahidol University,Bangkok 10400. Thailand

Key Words:

A study to responses of bird assemblages to dry evergreen forests and abandoned clearing areas after resettlement of hilltribes villages was conducted in Thung Yai Naresuan Wildlife Sanctuary, a natural world heritage site. The objective was to determine the birds¡¯ responses to certain factors including sizes of old settlement areas, time relapsed after the abandonment ot the settlements, and seasons. The permanent line transect technique was used for the field survey.

One hundred and eighty-five bird species were recorded. The diversity indices such as Shannon¡¯s index and Hill¡¯s number 1 tended to be lower in the old settlement areas than in the dry evergreen forest sites. However, the species assemblage of the two sites was clearly different and only the bird species assemblage in the 12-year-old abadoned settlement areas had the same composition as those in dry evergreen forest. Birds groups can be divided according to their species assemblages, plant successional stages, and seasonal change into 5 groups: opening species, secondary forest non resident species, dry evergreen forest non resident species, dry evergreen resident species, and generalist species. The abandoned settlement areas played an important role as feeding sites to the arboreal insectivore/frugivore and terrestrial insectivore bird groups. On the other hand, the dry evergreen forest played an important role as the feeding sites to the bark-gleaning insectivore, arboreal faunivore/frugivore, and foliage-gleaning insectivore bird groups. The study also suggested that the plant and the bird communities showed a clear recovery pattern after the resettlement of human beings.

No.16

Forest Industry Health in the Conservation and Management of Fragmented Forest Landscapes: A Case Study in Northern New England (USA)

Theodore E. Howard, Difei Zhang and Meghan Cornelison

Department of Natural Resources, University of New Hampshire Durham, New Hampshire, USA

Key Words: Forest industry, land rent, economics, northeastern US

The Ricardian model of land use, based on relative economic values of alternative land uses, suggests that the success of conservation and management of fragmented forest landscapes greatly depends on timber income. Timber income, in turn, depends on the health of the regional forest products industry. We conducted a survey of all sawmills in northern New England, USA, to profile work force characteristics, capital investment trends, log supply and final product market conditions, competition with neighboring Canadian mills, and the business environment in which the New England mills operate. Although responses often differed by mill production size class or location of mill, most respondents were concerned about the availability and cost of logs. Decision-makers should be careful that policies intended to protect forest land from conversion to other uses do not increase raw material cost and availability and thereby weaken the local forest industry.

No.17

Conservation of Biodiversity in Agricultural and Low Mountain Landscape

- Habitat Evaluation Using Raptors as Index of Diversity.

Hiroshi Momose

Chief Researcher, Wildlife Management Lab., National Agricultural Research Center

Key Words: Habitat selection, Modeling, Raptors

Two species of raptors, Goshawk (Accipiter gentilis) and Gray-faced Buzzard-Eagle (Butastur indicus), were studied in Ibaraki prefecture, about 90 kilometers north of Tokyo. The study area of 384 square kilometers contained several different land use and vegetation such as urban, suburban, rural, low mountain, and mountain areas. We found 30 Goshawk and 139 Buzzard-Eagle breeding pairs in this area. The goshawks were found in all areas except the urban area (Utsunomiya city), but Buzzard-Eagles were abundant only in the low mountain area, which was mainly composed of secondary forests and many branches of the valleys used as paddy fields. We constructed habitat selection models of the two raptor species to predict the breeding density within a 3 by 3 (Goshawk) and 2 by 2 (Buzzard) kilometer mesh using many environmental measures, such as vegetation, elevation, population etc., as independent variables. The models explained the breeding density of the two species fairly well. The largest positive factor chosen by the model was the length of forest edge facing grassland (Gosh) and the length of forest edge facing paddy field (Buzz). The largest negative factor was human population in both species. It was suggested that edge preference in the Goshawk made this species suitable for living in suburban, rural, and low mountain areas in this part of Japan, where forests are being fragmented because of sprawling type of suburban development. It was also suggested that the Gray-faced Buzzard-Eagle might be more vulnerable of the two raptors since paddy fields in the low mountain valleys are diminishing rapidly because of declining agricultural practice in this area.

No.18

A study of forest landscapes-a simple model of our efforts to make a symbiosis

in nature

Shigeo Nishizawa

The Japan Forest Engineering Society

Key Words: simple model, forest, landscape, effort, necessity, symbiosis

This is a study that briefly illustrates the relationship between the theories of landscape ecology and their application to managing forest resources. We think that by simplifying the environmental factors, our efforts to make a symbiosis in nature will be easy, not difficult.

We can think of the inside of the triangle as below having one apex of (fragmented) forests, and the other apexes of theory and its application. We, therefore, can think of its inside as all of the world using technology. Correspondingly, we can think of the outside of the triangle as having apexes of nature, science and society. We also can think of the circle as a friendly symbiosis between nature and society (science).

In a forest, our efforts to make a symbiosis are to think about (while imagining the diagram below) the following points:

¢Ù Science and symbiosis in a theory of landscape ecology. In short, it is to think scientifically about whether there are any factors (gap) preventing a sustainable symbiosis in the ecosystem, or not. It will, for instance, point out to us the necessity of an investigation of any influences on the environment.

¢Ú Society and symbiosis in an application to managing forest resources. In short, it is to think, as a society, about whether there is too much (gap) consumption of nature because of, for example, the demand by the economy for fuel, or not. It will, for instance, point out to us the necessity of making evident the over utilization of the environment.

Thinking of this simple model, our efforts to make a symbiosis in nature will produce the necessity to value the environment. This will make us enlarge our sensitivity or emotional attachment to the forest landscape beyond just technology because of the existence of our consciousness of beauty. It will make all citizens be able to understand better the causes of fragmented forests and make us make more efforts to make a symbiosis in nature.

No.19

Dynamics of Land-use/cover changes in Nairobi City

Charles Ndegwa Mundia¹ and Masamu Aniya²

1.Graduate School of life & Environmental Sciences, University of Tsukuba, Ibaraki 305-0005, Japan.
2. Institute of Geosciences, University of Tsukuba, Ibaraki 305-8571, Japan

Key Words: land use/cover changes, unsupervised classification, post-classification analysis, GIS

We used three Landsat images in a post classification analysis to map the land use/cover distribution for Nairobi area, Kenya. An unsupervised classification approach which uses minimum spectral distance to assign pixels to clusters was used. The overall accuracy of the three output images ranged from 87% ・90%, while the kappa coefficients ranged from 0.81 - 0.86. Land-use/cover statistics, extracted from Landsat Multispectral Scanner (MSS), Thematic Mapper (TM) and Enhanced Thematic Mapper plus (ETM+) images for 1976, 1988 and 2000 respectively, revealed that substantial land use/cover changes have taken place and that the built up areas have expanded by about 47 km2 over the study period. Forests have decreased substantially while agricultural lands have been on the increase. Parts of the formerly agricultural areas, forest areas and bush lands have given way to urban sprawl.

As Nairobi area has been developing rapidly over the last two decades, successful planning will require reliable information about land use/cover distribution and change. This study has demonstrated the usefulness of satellite remote sensing, digital image processing, and GIS techniques in producing accurate land use/cover maps and change statistics for Nairobi area for the last 24 years.

No.20

Spatial Relationship between Spaces of Forest comprises Shinto Shrines and Topographical Parameters in Urban area based on principles of landscape ecology

Naoko Fujita and Yoichi Kumagai

Graduate School of Frontier Sciences, The University of Tokyo

Key Words: Digital Map£¬ Urban£¬ Green Space£¬ Space of Forest comprises Shinto Shrines£¬ Digital Elevation Model£¬ religion

I focus on the landscape ecological evaluations of the spaces of Shinto shrines in urban area by the spatial-temporal scale approach. The space of forests, trees, mountains and seas, comprises faith in Shinto in origin. Nature vicinity of shrines should have been ustained. Despite the fact, the spaces of the Shinto shrines have been evaluated for religious field only. However, assessing the impact of land-use change is an important task in the context of global-change scenarios. Land-use was seen as an integrative variable, witch depends on natural, as well as on socioeconomic parameters. The spaces of the Shinto shrine are under the necessity of find out the new natural scientific worth. Planning for sustainable land management has become a very complex process. In this state, pay attention to the movement of 100 years recently is effective in scientifically catching the spaces of them. To evaluate the green spaces, it is necessary to presents values from different scales. A

nalyze geographical features and number of Shinto shrines, analyze distance of the central area of Tokyo and numbers of Shinto shrine, and analyze the Shinto shrine forest on various scales. This paper explores the possibility of using 5m-DEM (Digital Elevation Model) from digital map data base to analysis spaces of forest comprises Shinto Shrines and topographical parameters. 5m-DEM map enable to clear the altitude relief in this site. I observed that over thirty percent spaces of Shinto Shrine are located on ecotone; the transition area from watersides to lands or the slant area.

No.21

Occurrence and landscape ecology of a rare disjunct maple species:

Acer pycnanthum

Ikuyo Saeki¹ amd Burton V. Barnes²

1.Laboratory of Landscape Ecology, Faculty of Agriculture, Tokyo University of Agriculture and Technology

2.School of Natural Resources and Environment, University of Michigan

Key Words: red maple, rare species, landscape ecology, conservation

Japanese red maple, Acer pycnanthum K. Koch, is an endangered maple species that has a close taxonomic relationship with Acer rubrum L. Whereas A. rubrum is one of the most widely distributed and abundant species in eastern North America, A. pycnanthum is rare in central Honshu, Japan. Despite its endangered status, its landscape ecology is not well understood. Our objectives were to locate all known populations, characterize ecosystems where they occur, determine natural history characteristics that account for its persistence and restricted occurrence, and discover new populations by aerial survey using a landscape ecosystem approach. The 31 known, remnant A. pycnanthum-dominated ecosystems (< 18 ha) are confined to lower slopes of three river basins where wetlands have persisted and recurrently formed for at least 8 million years due to unique geological, topographic, and soil features. A. pycnanthum, an obligate wetland species, occurs in diverse, acid swamps, including nine seepage and four floodplain ecosystem types. These ecosystems are characterized by strikingly high species richness in the understory (up to 24 species per 600 m2) and groundflora (e.g., 48 vascular species per 25 m2), low similarities (< 37%) for both understory and groundflora species within and among ecosystems, and associated endangered species. Regeneration is restricted to sites with high light irradiance following disturbance. Because of the rarity of wetland sites, its inability to colonize upland sites occupied by dense natural vegetation or planted conifers, and the encroachment of agriculture and urbanization, A. pycnanthum is increasingly spatially limited. Aerial survey to date revealed 16 new populations, many of which are relatively large, intact ecosystems. This discovery provides a new focus for the conservation throughout the full range of wetland ecosystems as well as A. pycnanthum itself and its diverse and endangered associates.

No.22

Seasonal variation of transpiration in desert riparian cottonwood forest patches

Rico M. Gazal¹, R.L. Scott², D.C. Goodrich² and D.G Williams¹

1.Department of Renewable Resources, University of Wyoming, PO Box 3354, Laramie, WY 82071, USA

2.Southwest Watershed Research Center, USDA-ARS, Tucson, AZ 85719-1596, USA

Key Words: Populus fremontii, riparian vegetation water use, depth of groundwater, stand transpiration, sap flow

Patches of cottonwood (Populus spp.) forests occur along the floodplains of western United States. This riparian vegetation is dependent on a stable source of moisture throughout the season. Understanding the dynamics of seasonal transpiration of these systems in response to different environmental factors is necessary for determining riparian water balance and the amount of water required to sustain forest structure and function. Rates of transpiration (E) were evaluated in desert riparian cottonwood (Populus fremontii) patches at perennial and intermittent reaches of the San Pedro River in southeastern Arizona. Sap flow was measured using thermal dissipation probes and scaled to the stand level to investigate E in relation to canopy structure, depth to groundwater (GW) and vapor pressure deficit (D). The cottonwood stand located at the perennial stream site had higher leaf area-to-sapwood area ratio (0.31 ¡À 0.04 m2 cm-2), leaf area index (2.75) and shallower GW (1.1 to 1.8 m) than the stand at the intermittent stream site (0.21 ¡À 0.04 m2 cm-2, 1.75 and 3.1 to 3.9 m, respectively). Moreover, total annual E was higher at the perennial stream site (966 mm) than at the intermittent stream site (484 mm). During the peak dry period prior to the monsoon season, the cottonwood stand at the intermittent stream site exhibited high stomatal resistance in response to high D and low GW. However, E increased after significant monsoonal rains and runoff events that recharged groundwater at both sites. Riparian cottonwood forests exist across a continuum of groundwater availability and their access to shallow groundwater sources determines structural and physiological responses to drought. Increases in water table depth will enhance the susceptibility of these forests to drought stress, which may threaten their productivity and existence along river systems in arid environments.

No.23

Seasonal variation of edge effects on the microclimate, vegetation and light environment of primary, secondary and urban forest fragments

in southern Hyogo Prefecture

Hiroaki Ishii and Ayako Iwasaki

Graduate School of Science and Technology, Kobe University

Key Words: biodiversity, forest conservation, hemispherical photographs, lucidphyllous forest, shrine, temple

In Japan, forests associated with shrines and temples represent remnant areas of native vegetation in a human-dominated landscape. We compared the extent of edge effects on the vegetation and light environment among primary, secondary and artificial shrine/temple forests in Hyogo Prefecture: lucidophyllous primary forest and deciduous secondary forest at Taisanji Temple (16-ha natural forest) in western Kobe City surrounded by agricultural fields and artificial laurel forest at Nishinomiya Shrine (4-ha urban forest) surrounded by buildings and roads. Hemispherical photographs were taken at 1-m distance intervals along a 40-m transect to characterize light environment with increasing distance from forest edge. Diameter and height were measuremed for all trees within 20 m of the transect. In the primary forest deciduous trees were found near the edge, whereas Castanopsis cuspidata dominated the canopy of the interior. Canopy openness decreased markedly from the edge to about 10 m into the forest and stabilized at 9-12% in the forest interior. In the secondary forest, various tree species dominated the canopy from the edge to the interior. Canopy openness decreased from the edge to about 20 m and stabilized at 10-17% in the interior. Edge effects on the light environment were observed for distances 1.5-1.7 and 1.3-3.0 times the canopy height in the primary and secondary forest, respectively, suggesting that edge effects penetrated deeper in the secondary forest. In the urban forest, non-forest species such as Pseudosasa japonica, and Trachycarpus excelsa. were found throughout the forest. The forest edge was surrounded by a concrete wall and light environment did not show directional change from edge to interior. This suggested that human intervention such as removing non-forest species and planting late-seral species may be necessary to create natural-forest-like conditions in urban forests

No.24

Plant species occurrences in a Larix kaempferi plantation dominated landscape

in central Japan: effects of stand age

Takuo Nagaike, Atsuko Hayashi, Masako Kubo, Midori Abe and Nobumasa Arai

Yamanashi Forest Research Institute, Masuho, Yamanashi, 400-0502, Japan

Key Words: Abundance of plant species, Life form, Long-rotation plantation, Plant species diversity, Seed dispersal type

In Japan, most of forest landscapes are formed by mosaic with small patches, because of long history of forest management. Moreover, plantation in Japan is major forest type and comprised 40% of forest area. In order to ecologically manage in the landscapes, we cannot ignore the plantation and should consider to age distribution of plantation. In central Japan where Larix kaempferi plantation and abandoned coppice forest are major landscape elements, we clarified the plant species diversity of forest-floor vegetation in the two forest types and the effects of stand age on plant species diversity in plantations. Both plant species richness and abundance in younger plantation are higher than those in older plantation, although young plantation contained much weed and disturbance-tolerant species. On the other hand, species composition in older plantation was similar with that in abandoned coppice forest. We will discuss about ecological management, especially age distribution of the plantation, in the landscape.

No.25

Extraction of stepping-stone corridors for birds in urban areas

using remote sensing and GIS

Hiroshi Hashimoto¹, Jianjun Dong¹, Junichi Imanishi² and Yukihiro Morimoto²

1.Graduate School of Agriculture, Kyoto University

2.Graduate School for Global Environmental Studies, Kyoto University

Key Words: ecological network, birds, isolated woods, stepping-stone corridor,remote sensing, GIS

Some residential birds widen their habitat ranges in non-breeding season. Their wintering habitats may be determined not only by quality of habitats but also by connectivity from source habitats. In this study, we extract some possible stepping-stone corridors, and evaluate which corridors explain presence / absence of four residential birds: Bush Warbler, Long-tailed Tit, Varied Tit, Masked Grosbeak, in possible wintering habitats in Kyoto City. Bird survey was conducted in 18 woods (> 0.6 ha): almost all possible breeding patches for these four species in the urban area of Kyoto. We assume patches in which each bird species was recorded during breeding season and mountain forests around the city as their source habitats, and analyze whether their source and possible wintering habitats are connected by corridors. A 15 m by 15 m resolution of vegetation cover ratio (VCR) map was derived from Terra / ASTER and Quick Bird images. We assume minimum VCR for stepping-stones as 0.25, and create buffer zones from each stepping-stone in four distances: 50 m, 75 m, 100 m and 125 m, then four types of corridors were extracted by continued buffer zones. Note that the maximum interval between stepping-stones is twice the distance of the created buffer. The maximum interval for suitable corridor evaluated by maximum Cohen's kappa, maximum overall prediction success and minimum interval that more than 80% of actual wintering habitats connect to the source habitats were, respectively, 200m (kappa 0.25), 200-250 m (75% success) and 200m (89% connection) for Bush Warbler, 100 m (0.12, 58%) and 150 m (80%) for Long-tailed Tit, 100 m (1.0, 100%, 100%) for Varied Tit, 100 m (0.29, 64%) and 250 m (100%) for Masked Grosbeak. Though further analysis may be necessary, this result gives us some suggestions about the interval of stepping-stones for planning ecological corridors.

No.26

Landscape dynamics and avian adaptation to forest fragmentation caused by natural fires in British Columbia, Canada

Kenneth R. Parker

Proprietor, Birdland Ecosystem Management, Prince George, British Columbia CANADA

AIRIES Visiting scholar, National Institute for Environmental Studies (NIES), 16 ¨C 2 Onogawa, Tsukuba, Ibaraki 305-0053, JAPAN

Key Words:

Better emulation of landscape patterns associated with natural disturbance, such as fire or insect epidemic, could go a long way toward improving the conservation of biodiversity in forests that are managed for timber. I examined this concept in British Columbia and Alberta, CANADA by sampling bird and mammal communities in natural landscapes containing patches of older forest that differed in size and mechanism of origin. Much of my work focused upon coniferous forests, notably those dominated by seral lodgepole pine, where we examined the landscape patterns associated with large historic burns. Small patches in these forests, called remnant patches (RP), were created within the boundary of a burn. RP ranged from 0.030 to 20 ha (mean, 6.5 ha; median 4.4 ha) comprising 10 ¨C 15 % of the burn area of a large fire. Some burns contained comparatively few patches. Larger patch sizes of old-growth forest (OG) with similar coniferous canopies (mean age, 200 years) that originated through exclusion between different fires were compared with the smaller patches inside burns. Based on point counts and transects, bird abundance was higher in OG than RP, and both were higher than the regenerating matrix forest (SM) of the former burn. Barring spurious observations of rare or displaced species, there were no species to speak of in OG that were not also represented in RP. Riparian habitats within RP, OG and SM had more species than their upland counterparts. Upland RP larger than ~6 ha contained as many species per unit of search as the larger OG forests. Based on my studies of natural burns in coniferous forests in central British Columbia and both harvested areas and natural burns in Alberta, I offer an approach toward designing these features in managed forests. Planning larger areas of harvest commensurate with the area of a natural fire would offer better perspective of scale to incorporate factors such as patch density, size and shape, hydrology, and variety in composition among a cluster of patches. Extending the idea that variety of habitat features improves biodiversity of a larger landscape as a whole, there are many habitat features apart from old-growth forest that may need to be considered in the overall plan. A few examples will illustrate this idea. I will discuss the influence upon bird populations of modeling landscapes after natural disturbance in relation to forest fragmentation, metapopulation dynamics and regulation of insect epidemics.

No.27

Bird communities in a fine-grained heterogeneous landscape in Japan

Yuichi Yamaura, Kazuhiro Katoh, Go Fujita and Hiroyoshi Higuchi

Graduate School of Agricultural and Life Sciences, The University of Tokyo

Key Words: birds, landscape structure, heterogeneous landscape, plantation forest, stand structure

There were few studies comparing the relative importance of patch-level variables with that of landscape-level variables for wildlife in less-contrasted landscape such as forested landscapes, so the importance of landscape perspective in landscapes other than agricultural landscapes is unclear. We examined the influences of stand-level and landscape-level variables on bird communities in a Japanese rural landscape. Japanese rural landscapes provide appropriate study fields to address this problem because forests in the landscapes consist of heterogeneous habitats for birds, namely, short coppices due to clear-cutting, coppices with sparse understory layer due to gathering litters, abandoned tall coppices with dense understory layer, and plantation forests. In addition, we also examined whether plantation forests could function as habitats for birds.

The study area is almost 3.5 by 2.5 km in extent, located in Tochigi prefecture, central Japan. Birds in 25 coppice stands and 18 plantation forest stands were censused six times using line transect method from late January to February, 2001. Bird species were assigned to the following five guilds based on their autoecology: litter searcher, bush user, Paridae group, and edge species. The relationships between the presence of each guild and environmental variables were explored by using logistic regression analysis.

The relative importance of stand-level variables and that of landscape-level variables were guild-specific. The Paridae group, seedeaters, and edge species were largely influenced by landscape-level variables, while litter searchers and bush users were associated with stand-level variables. The differences in their responses to landscape-level variables were considered to be related to food resource availabilities. Although the effect of forest-type was the largest for the Paridae group, its effect on other guilds was less important. It is suggested that plantation forests with dense understory layer could function as habitats for the litter searchers and bush users.

No.28

Forest bird community in fragmented old-growth natural forest-Effects of forest management practice on bird community in warm-temperate forest in Japan

Shigeho Sato

Shikoku Research Center, Forestry and Forest Products Research Institute

Key Words: bird biodiversity; confer plantation; forest utilization; human impact

Old-growth natural forest is extremely fragmented to small remnants because of forest utilization in warm-temperate zone of Shikoku district, Japan. In this area, most of natural forest had been replaced to copse and conifer plantation. In order to elucidate the impacts on forest bird community by forest utilization, the author compared the species composition of forest bird community in 20 research plots in Shimanto river basin, southwestern part of Shikoku district. The research plots were set in 8 old-growth natural forests, 5 secondary forests and 7 confer plantations. The number of bird species of old-growth natural forests was not significantly higher than those of secondary forest or of confer plantation. But species composition of old-growth forests was distinguished from those of secondary forests and of conifer plantations. More tree trunk users and tree cave users, which need dead trees or large trees with caves, were recorded in old-growth natural forest. Also some species which prefer dense understory inhabited mostly in old-growth forests. Those species include many rare species. It was suggested that forest utilization had limited the habitat area of those species.

No.29

Effects of landscape change on wildlife in Malaysia: A case study in Pasoh Forest Reserve and adjacent forest fragments in Negri Sembilan, Peninsular Malaysia.

Shinya Numata¹, Masatoshi Yasuda², Sen Nishimura^3,5, Keiichiro Yoshida⁴, Toshinori Okuda¹, Eng Seng Quah⁵, Mazlan Hashim⁶, and

Nur Supardi Md Noor⁵

1 National Institute for Environmental Studies, Tsukuba, Ibaraki 305-8506 Japan; 2 Forestry and Forest Products Research Institute, Tsukuba, Ibaraki 305-8687, Japan; 3 Japan Wildlife Research Center, Taito-ku, Tokyo 110-8676, Japan; 4 Yokohama National University, Yokohama-ku, Kanagawa 240-8501, Japan; 5 Forest Research Institute Malaysia, Kepong, Kuala Lumpur 52109, Malaysia; 6 Universiti Teknologi Malaysia, Skudai, Johor 81310, Malaysia

Key Words: Human disturbance, Landscape, Pasoh Forest Reserve, Peninsular Malaysia, Wildlife

Species richness and biodiversity are the most distinctive features of the tropical rain forest in Southeast Asia. Recently, however, these forests have been disturbed on a large scale through human activities, resulting in fragmentation and internal disturbance of the remaining forests. In order to conserve biodiversity of tropical rain forests we need to understand the past, current and future human impacts on biodiversity. Such studies should focus at a landscape level to include forests, agricultural plantations, and urban areas. Here, we choose an area which includes primary and secondary forest patches, agricultural fields, pastures, clearcuts, silvicultural plantations, and residential areas. Firstly, we examined human disturbances which occurred in the focal area. Changes in landuse patterns in the study area have produced rapid decline of total forest area from the 1970¡¯s to 1996; consequently the Pasoh Forest Reserve centered in the study area has become isolated. Secondly, in order to determine the potential species composition of a focal forest reserve in the study area, we compiled the results of six published studies and one Ph. D. thesis. These reports suggest that 111 species belonging to 11 orders were considered to be potential residents of the Pasoh Forest Reserve at one time. Thirdly, we conducted camera-trapping in the Pasoh Forest Reserve and adjacent fragmented forests to understand actual species composition of mammals. The camera traps detected 42% of non-flying mammals (17 species) in the forest. Invasion of domestic animals and poaching remain high in the reserve. Finally, we monitored road-kills of animals around Pasoh Forest Reserve and determined a high amount of road-kill impacts to middle-sized and migrant animals.

No.30

From microhabitat selection to landscape management of the Japanese marten and the Japanese squirrel

Midori Saeki¹, Hitoho Yatake², Tsuyoshi Akita², Jun Furukawa², Yasuo Iizuka¹, and Nobuo Fujiwara¹

1.Landscape and Ecology Division, Environment Department, National Institute for Land and Infrastructure Management, Ministry of Land, Infrastructure and Transport, 1 Asahi, Tsukuba-city, Ibaraki 305-0804 JAPAN

2.Environmental Research Department, CERES, Inc. (Civil Engineering Research & Environmental Studies), 1646 Abiko, Abiko-city, Chiba 270-1166 JAPAN

Key Words: marten, squirrel, habitat selection, landscape management

The Japanese marten (Martes melampus) and the Japanese squirrel (Sciurus lis) are native, arboreal species in Japan. Habitat selection of the two species was analyzed using the GIS database constructed for the study in Ibaraki Prefecture by the National Institute for Land and Infrastructure Management.

Habitat preference was obtained from field signs of martens, comparing with random samples generated via GIS tools. The marten signs were found only in forest habitat types, and they used more deciduous and red pine forests than expected, while showing no preference for coniferous plantation. The distances to roads and rivers from marten signs were significantly greater than expected.

All stands >10 ha were surveyed for presence of squirrels, vegetation type, forest height, canopy closure, and topography. Squirrels were present in 113 stands of 277 stands surveyed, mostly in the western mountainous parts of study area. All stands with squirrels had ¡Ý41% of canopy closure with the highest frequency between 71 to 90%. Evergreen softwood stands were commonly used, as well as mixed stands. There were strong associations between squirrel inhabitance and presences of red pine, Japanese cedar, Japanese cypress, and Japanese walnut. Also, two isolated stands were monitored, and diminishing habitat for squirrels was detected. Radio-tracking revealed high use of red pine trees by the squirrels.

Both forest fragmentation in mountainous areas and forest isolation in plains can be serious problems for martens and squirrels; especially for the squirrels, isolated populations have been disappearing. Our results suggested that habitat management for squirrels consists of tree species composition, canopy closure, and habitat connectivity, while martens appeared to be able to utilize mixed stands on ridges in plantation areas that might provide both habitat connectivity and food.

No.31

Sustainable management of fragmented forest landscapes via agroforestry innovations: Documentation of farmers' practices through PRA, household survey and computer database

Damasa M. Macandog^1a, Robert G. Visco^1b, Bobby A. Crisostomo², Marc M. Delgado^1c and Princess Alma B. Ani^1c

1a.Assistant Professor, Institute of Biological Sciences; 1b.Assistant Professor, Institute of Renewable and Natural Resources; 1c.Research Assistant, Institute of Biological Sciences, University of the Philippines Los Ba¬Ðs, College, Laguna, Philippines

2.Chief, Database Management, National Mapping and Resource Information Authority, Makati, Philippines

Key Words: agroforestry, documentation, PRA, household survey, computer database

Agroforestry systems are considered as sustainable land-management options in upland areas because of their potential ecological, economic and social benefits. Documenting the human impacts and landscape dynamics of agroforestry systems in the uplands is a challenge to researchers because of the enormous data to be collected and the complexities involved in gathering the required information. A method to document the innovations in the agroforestry systems adopted by farmers in Claveria, Misamis Oriental, Philippines, is presented using a combination of Participatory Rural Appraisal (PRA), household survey techniques and computer database.

Participatory rural appraisal (PRA) activities were conducted among selected farmers to provide a preliminary assessment of the site, identify research gaps and needs, and familiarize the researchers with the community. These include (i) transect mapping to identify various type of agroforestry systems practiced; (ii) timeline to capture the history of tree domestication in the area; and (iii) mind mapping to understand the motivations of farmers to adopt agroforestry. In the household survey, a total of 300 households were interviewed representing farmers practicing various agroforestry systems in different elevation classes. Farmers' demographic characteristics, farm biophysical resources, household socioeconomic data, motivations for planting trees, and ecological knowledge were recorded. The information gathered in the survey is managed and maintained in a database written in MS Access to facilitate data entry and retrieval.

Innovative agroforestry systems documented using PRA and household survey activities are described. The survey database developed in this study is introduced and its uses are provided, including linkage to GIS and to spreadsheet software for additional analyses. The techniques employed, coupled with the database will help advance the management of uplands through a powerful alternative in documenting agroforestry practices. Knowledge generated using the approach can assist in agroforestry planning, research and extension.

No.P-1

Forest landscape changes around Ogawa Forest Reserve, Ibaraki, Japan,

based on old topographic maps and aerial photographs

Asako Miyamoto and Makoto Sano

Department of Forest Management, Forestry and Forest Products Research Institute

Key Words: landscape change, human impact, coniferous plantation, grassland

It is important to comprehend the history of the environment, such as the land use and disturbance, because it provides us valuable insight into current ecosystems and biodiversity. In order to obtain the knowledge for proper conservation and management decisions, we reconstructed the past forest landscape around Ogawa Forest Reserve, which was established in 1969 in northern Ibaraki, Japan, using old topographic maps and aerial photographs, and identified the changes that had occurred from the 1900s to the 1990s. The study area was classified into four categories: grasslands, secondary deciduous broadleaf forests (secondary forests), coniferous plantations and non-forested lands. Forest landscape patterns were quantified by several landscape indices related to their area, density and shape to understand their spatial distribution and connectivity. The results showed that the forest landscape has been drastically changed by human impacts, from grasslands and secondary forests to fragmented secondary forests and coniferous plantations. Grasslands have been largely removed from the landscape, with only small fragmented patches remaining. Secondary forests have also decreased in area, becoming patchier and more irregular in shape. In contrast, coniferous plantations have increased due to extensive afforestation of sites which were once occupied by secondary forests. Overall, forest landscape has become fragmented and complicated during the past 90 years.

No.P-2

Predicting a habitat suitability of the alien Formosan squirrel

in Kanagawa Prefecture, central Japan

Asako Miyamoto¹ and Noriko Tamura²

1.Department of Forest Management, Forestry and Forest Products Research Institute

2.Tama Forest Science Garden, Forestry and Forest Products Research Institute

Key Words: alien species, Formosan squirrel, logistic regression model, Geographic Information System (GIS), potential breeding site

Alien species damage native species and ecosystems on a global scale. Thus, to prevent further loss of biological diversity, we need to study and control the spread of alien species as quickly as possible. The Formosan squirrel, Callosciurus erythraeus thaiwanensis, is an arboreal mammal that originated in Taiwan. This species was first naturalized in Japan after it escaped from the zoological garden on Izu-Oshima, Tokyo, in the 1930s. After that, it became established among several sites such as Kamakura City, Wakayama Castle Park and Himeji Castle Park.

In the present study, we have clarified the environmental variables that influence the presence of the Formosan squirrel and made a logistic regression model using those variables in Kanagawa Prefecture. Furthermore, we have compared the results of a field survey with the predictions of the squirrel's occurrence and identified areas in which stronger countermeasures are required to prevent expansion of the squirrel's range.

As a result, we found that woods suitable for the Formosan squirrel were distributed not only around Kamakura, which Formosan squirrels currently inhabit, but also throughout Kanagawa Prefecture. Furthermore it was apparent that the area west of the Sagami River is a large continuous site suitable as Formosan squirrel habitat. To prevent intrusion west of the Sagami River, it is important to monitor the areas, in which many woods are concentrated and might be used as breeding habitats, such as woods along Sagami Bay and along the Sagami River, because it is possible that the Formosan squirrel will invade the western area via these woods.

No.P-3

Land use/cover change evaluation based on satellite imagery and GIS:

A case study in the Bindura district, Zimbabwe

Courage Kamusoko* and Masamu Aniya**

* Graduate school of Life and Environmental Sciences, University of Tsukuba, Tennodai 1-1-1, Tsukuba, Ibaraki 305-8752, Japan

**Institute of Geoscience, University of Tsukuba, Tennodai 1-1-1, Tsukuba, Ibaraki 305-8751, Japan

Key Words: Land use/cover changes, hybrid supervised/unsupervised classification, GIS, post-classification comparison.

Land use/cover change evaluation using satellite imagery and GIS for the Bindura district in Zimbabwe, Africa is critical for sustainable environmental management. In light of the recent reports of widespread deforestation resulting from the fast track land redistribution programme, there is an urgent need to monitor land use/cover changes. Landsat 5 Thematic Mapper (TM) data of 9 September 1986 and 2 September 1995, and Landsat 7 Enhanced Thematic Mapper plus (ETM+) data of 9 October 2000 were used in the image classification. A hybrid supervised/unsupervised classification approach coupled with GIS analyses such as neighborhood and recode were employed to generate land use/cover maps for 1986, 1995 and 2000. Classification accuracies of the three maps are above 85%. A post-classification comparison change detection technique was performed on the land use/cover maps. The rate of land use/cover changes was observed to vary from 36% to 26% during the study period. The major land use/cover conversions were from woodland and mixed rangeland to agriculture for 1986-1995, and from woodland and agriculture to mixed rangeland for 1995-2000. Observed trends in land use/cover changes can be partly explained by reforms in the agricultural sector in the early 1990s, the impact of droughts and the agrarian reforms initiated in 2000. Detailed quantitative information presented in this study provides useful insights into the complex patterns of land use/cover changes, which can be used for evaluating environmental problems such as soil erosion and woodland degradation within the study area.

No.P-4

Lantana camara and the changing landscape of forest ecosystems in India

D.R. Batish, H.P. Singh and R.K.Kohli

Centre for Environment, Department of Botany, Panjab University, Chandigarh 160 014, India.

Key Words: Lantana camara, Exotic Invasive weed, Allelopathy, Shiwalik Himalayas

Lantana camara L. (family Verbenaceae) is an exotic invasive weed from Tropical America that has invaded the natural forest areas, plantations, wastelands and grasslands of India. The weed because of its dense shrubby growth invades and occupies the available niches in these areas and soon overtakes the natural vegetation. Besides, the competitive and allelopathic ability of the weed also contribute to its fast spread and encroachment of areas. In addition, L. camara also adversely affects the natural regeneration of tree species and thus changes the community structure and landscape. In the forest areas and along their boundaries presence of thickets of this weed has become the predominant feature. The situation is worst in the Shivaliks ¨C a range of lower Himalayas in Northern parts of India where hill-tops have been fully covered by this weed replacing the naturally growing indigenous trees. Further, because of the presence of pungent active principles including also the volatile essential oils in this weed, it is not relished by the animals and thus leads to fodder scarcity. Our studies in the Shivaliks have shown that L. camara interferes with the growth and development of natural vegetation through allelopathic mechanism besides resource competition. Due to its allelochemistry, it also alters the status of soil nutrients thereby limiting their availability for the growing vegetation. The present study discusses all these aspects of the interference mechanism of L. camara with the natural vegetation of Shivaliks leading to a change in landscape.

No.P-5

Quantitative analysis of forest fragmentation in Patagonia, Argentina

Francisco Andres Carabelli

National Council for Scientific and Technical Research of Argentina - Patagonian Andes Forest Research and Extension Center

Key Words: Patagonia, native forest fragmentation, landscape change

Two main central issues in the study of forest fragmentation in Patagonia related to human-induced alterations are nowadays of great relevance: changes at landscape level across time and a better understanding of the fragmentation process at a landownership level. To improve the comprehension of the first problem, we selected one of the native tree species more severely affected by fragmentation processes. The Austrocedrus chilensis (cipres de la cordillera) has been historically affected by unplanned development of human settlements increasing risk of fires, irrational use for cattle grazing, replacement by exotic conifer species and intensive timber exploitation. We analyzed two alterations caused by forest fires and replacement by exotic conifer species in one emblematic area of `cipres´ geographical distribution representing 2,5% of the total area covered by this species in the Patagonian Andes of Argentina, to quantify changes in landscape heterogeneity between 1970 and 2001. The results showed that the areas dominated by `cipres´ forests forming a continuous or interconnected area of almost 3400 hectares in 1970 had been drastically modified in 2001. A net area reduction of 24% was accompanied by a strong negative change in forest landscape heterogeneity due to fragmentation of `cipres´ forests; almost 34% in the considered period of time. Significant reductions were checked in the average patch size; 35 hectares in 1970 while 10 hectares in 2001-. The Biggest `cipres´ patch occupied in 1970 comprises an area of almost 500 hectares, whereas in 2001 almost 60% smaller at about 210 hectares. These results become relevant if we consider the multifaceted fragmentation process has an increasing influence on the degradation of environments and biodiversity decline. Hence, quantitative evaluation of forest fragmentation has become essential to support a decision-making practice leading to a more efficient protection of these unique ecosystems.

No.P-6

Spatial distribution change of endangered Amami rabbit Pentalagus furnessi after invasion of small Indian mongoose Herpestes javanicus on Amami-Ohshima, Japan

Fumio Yamada¹, Ken Sugimura¹ and Shintaro Abe²

1.Forestry and Forest Products Research Institute (FFPRI)£¬PO Box 16, Tsukuba-Norin, Ibaraki 305-8687, Japan¡¡E-mail: fumio@ffpri.affrc.go.jp

2.Amami Wildlife Conservation Center, Yamato 894-3104, Kagoshima, Japan

Key Words:

The exotic small Indian mongoose Herpestes javanicus (Family Herpestidae, Order Carnivore, Mammalia) was introduced for pest control of snakes and rats on Amami-Ohshima Island in 1979, which has many endemic and threatened species and is located in the Ryukyu Archipelago in the southwestern most part of Japan. However, the mongoose has caused a great damage on crop production and poultry in farmland and furthermore has had a predatory impact on the native animals living on the Island. So, the mongoose is one of the most urgent species to eradicate among invasive mammals in Japan. Recently, the Ministry of the Environment of Japan has started full-scale control of the mongoose from 2000 as a model for conservation of biodiversity of the subtropical islands.

For monitoring of invasive mongooses and of impacts of native species, we investigated the animal fauna by auto censor camera and the change of distribution by fecal pellet counts census of Amami rabbit Pentalagus furnessi, which is one of the most flagship species on Amami-Ohshima Island. The surveys of auto camera were conducted during the period from November 2000 to March 2004. Total 30 auto cameras were set at the degree of different density of mongoose for 1-2 weeks in each survey. The surveys of pellet counts census were conducted during the period from February 2002 to February 2003. Rabbit pellets were searched and counted along forest roads and streams as well as inside forests.

The frequency of the number of photographs of mongoose to total those of animals in the northern area (Naze City and Yamato Village) was high (10%), but low (1-2%) in the southern area (Sumiyo Village) where rabbit density is high. Mongooses were also found by the cameras in the areas where trappings had been discontinued due to low incentive of trappers or prohibitions by landowner. In addition, our investigation also found that a mongoose entered into the breeding nest of a rabbit in daytime in the low mongoose density in the south. Furthermore, the rabbit population was found to be rapidly decreasing and disappeared in less than eight years in relatively high-density mongoose area, especially in the north.

These results indicate that if a small population of mongooses, even only one mongoose, remains in rabbit habitat, the rabbit will become extinct due to predation by mongooses not only on adult rabbits but also on juveniles inside the breeding nest. Therefore, it is necessary to eradicate mongooses in the habitat of the rabbit and to prevent of their invasion into other rabbit habitat.

No.P-7

Forest edge environment preferred by goshawks in Ishikari Plain, Japan

Gen Takao¹, Kenichi Ozaki¹ and Tomotake Sakai²

1.Hokkaido Research Center, Forestry and Forest Products Research Institute, Sapporo 062-8516

2.EFP, Co. Ltd., N18W3-21, Sapporo 001-0018

Key Words: goshawk, forest edge, home range, vegetation map, hypothetical point distribution, Hokkaido

Goshawk (Accipiter gentiles L.) is a raptor which prefers forest edges in a man-caused patchy landscape of forests and open lands. They have been observed in Ishikari Plain, Hokkaido, northern Japan, where small but long forest stands, e.g. windbreaks or riparian forests, remain among widely-opened arable lands. To measure the dependency of the goshawks in Ishikari Plain on the forest edges, distributions of the distance between forest edge and location point (edge distance) for the observed males were analyzed. The edge distance of a location point was calculated from the vegetation map which had been produced using Landsat TM images and aerial photos. Of whole males, 98% of the location points inside forest were dropped within 100m from edge, while 96% among outside forest dropped within 300m. The area outside and within 300m of the edge had a strong correlation with the home range area, but the area inside and within 100m did not. This suggested the forest area in a home range was independent from the home range size. For each male, the mean edge distance outside forest were compared and tested against the ones from some simulated hypothetical point distributions in the home range; one was a random (Poisson) distributed points, and another is an inversely proportionally distributed points from edges. The results revealed that majority of the males had significantly shorter distances than the random distribution but not significantly different from the inverse proportional, though a few males had even longer than the random.

No.P-8

GIS-Based CMA for Deforestation Risk Model in Tung Salaengluang National Park, Thailand

Jaruntorn Boonyanuphap

Faculty of Agriculture, KOCHI UNIVERSITY

Key Words: Deforestation, Risk Model, Geographic Information System, Tung Salaengluang

The GIS-based approach has been applied for monitoring the change of land use and forest area between year 2000 and 2002 in Tung Salaengluang National Park with a buffer zone at a radius of 2 kilometers, Phitsanulok, Thailand. The study focuses on integrating Geographic Information System (GIS) with Complete Mapping Analysis (CMA) to express the vulnerability value of risk factors in order to develop a deforestation risk model. The study area was classified into 7 major land use categories, namely, forest area, agricultural land, community area, grassland, water bodies, old clearing and shifting cultivation, and other minor categories. The forest area had increased approximately 7,856.32 square kilometer due to promotion of forest plantations. Forest areas have mainly been changed into agricultural land. Nine variables relating to a bio-physical environmental factor and a human activity factor was used to determine the deforestation risk model. Forest type and distance from a

gricultural land were the most important of all bio-physical environmental and human activity factors for deforestation, respectively. The human activity factor had more influence than the bio-physical environmental factor for the risk of deforestation. The study area fell into a range of five deforestation risk categories from very high risk (VH) to very low risk (VL). The existing forest area was mostly classified as the high risk class (H). The accuracy of the deforestation risk model had been evaluated using the coincided area between deforestation risk class and actual deforestation area in year 2000/2002. The very high risk class (VH) had the highest accuracy at 32.88 among the other classes. This research expands the basic function of GIS technology to map the deforestation risk zone at different severity levels, which could give effective information for developing deforestation prevention activities in studies areas.

No.P-9

Landscape resarch in the Chequamegon National Forest: an overview

Jiquan Chen and Lees Menbers Crow

Landscape Ecology & Ecosystem Science (LEES), Earth, Ecological and Environmental Sciences (EEES), Bowman-Oddy Laboratories, University of Toledo

Key Words: Chequamegon, disturbance, management, hypothesis, theories

The LEES group and the USDA Forest Service have conducted landscape research since 1994 in a managed landscape in northern Wisconsin. Our research focuses include landscape structure and dynamics, plant distribution, microclimate, CO2 and H2O fluxes, disturbances, and production, using various modeling, spatial statistics, remote sensing, and GIS approaches. I this paper, we provide an overview of the major research components and lessons learned from 10 years efforts. New knowledge includes scale-dependent relationships between structure and processes, significant contributions of different landscape elements, and difficulties in incorporating ecosystem processes into complex land mosaics. Six research topic areas are: (1) provide a broad-scale characterization of landscape change (1972-2001) using satellite imagery and detailed field data. A GIS-based D-AEI model and new methodology for delineating area of edge influences were developed. Quantitative methods (e.g., wavelet analysis) were used to examine the changes of landscape at multiple scales; (2) quantify and compare understory vegetation diversity (e.g., richness, abundance) among landscape elements such as patch type, edge, roadside, riparian zone, etc., (3) examine and predict the interactions between landscape structure and microclimate (e.g., temperature & moisture), soils, micro-topography, and disturbances; (4) examine the effects of harvesting and simulated fires on landscape structure and production using PnET, HARVEST, FORSITE, FVS, LandNEP, EV, RS products, USGS spatial database, and the management maps; (5) investigate the central hypothesis that the land mosaic (i.e., various ages & types) determines the cumulative NEP of disturbed landscapes. Direct measurements of carbon exchange (eddy-covariance towers), physiological variables, and ecosystem structure are combined with models of ecosystem processes and disturbances on land mosaics, and remote sensing imagery; and (6) synthesize our major discoveries, compile data and model results, publish models and RS products, and deliver results to both scientific and management communities.

No.P-10

Fires, Management, and Land Mosaic Interactions:

A Generic Spatial Model and Toolkit from Stand to Landscape Scales.

Jiquan Chen, T.T. Crow, S. Ryu, D. Zheng, J. LaCroix, and B. Song

Landscape Ecology & Ecosystem Science (LEES), Earth, Ecological and Environmental Sciences (EEES), Bowman-Oddy Laboratories, University of Toledo

Key Words:

To understand the accumulation of wildland fuels and fires and to develop sound new management guidelines, the most critical research needs are a synthesis and integration of our knowledge of fire ecology, ecosystem processes, and landscape ecology, with computer models and simple tools that allow the public, land managers, and other decision makers to assess fire risk and predict the influences of fire on a given landscape. Such tools should be easy to use, with options for manipulating key factors so that alternative scenarios can be examined in a cause-effect manner to support final management decisions. The primary objective of our study is to develop a spatially explicit, PC Windows-based generic model accompanied by visualization systems that land managers can readily access to examine the potential effects of fire regimes and forestry practices on the landscape mosaic, ecosystem dynamics, fuel load patterns, and fire risk in real landscapes. Specifically, our objectives are: (1) To incorporate the interactions among wildfires, forest management, and current and future landscape mosaics and dynamics in a simple, generic, PC Microsoft Windows-based tool for resource managers and the scientific community. A visualization module will be the central component of the model, allowing users to best apply their knowledge of fire management at multiple temporal and spatial scales and to see an image of the results for stand and landscape structure; (2) To analyze intensely the ecological consequences of fire and harvest in the Chequamegon landscape and to explore alternative fire and management scenarios for this landscape; (3) To test the cause and effect relationships among fire, management, and land mosaics at the landscape scale for six landscapes in addition to the Chequamegon which has distinct ecosystems, climates, fire regimes, and management guidelines; and (4) To organize workshops and promote the application and refinement of the model. Participants will include researchers, land managers, planners, and students. Six research Tasks have been identified by an interdisciplinary team with expertise in fire ecology, landscape ecology, modeling, and regional assessment to examine fire, management and their interactions with ecosystems at seven landscapes centered on National Forests across the country.

No.P-11

Sustainable land use management planning in the Phu Fa project, Nan watershed, Northern Thailand

Maliwan Tanasombat¹, Prakit Vongsrivattana² and Katsutoshi Sakurai

1.Laboratory of soil environmental science, Faculty of Agriculture, Kochi University

2.National Park, Wildlife and Plant Conservation Department, Bangkok, Thailand

3.Laboratory of soil environmental science, Faculty of Agriculture, Kochi University

Key Words: Sustainable, Land use, Management, Planning, Phu Fa project, Nan watershed

The Phu Fa project is located on the steep slopy mountainous areas of Bo Klua district, Nan province, which is a part of a Nan watershed area in Northern Thailand. The areas elevation is 500 - 700 m a.s.l. with maximum and minimum temperature at 35.9 in April and 3 in January, and the average relative humidity is 79.8 %. This area was classified as a 1A watershed area according to the watershed classification system which is used only in forest areas. In fact, more than 70 % of the whole area is used for agriculture, especially with upland rice without intensive management. This kind of fragmentation causes serious erosion and land slides.

In this study the existing land used map was investigated. This area was classified into 5 categories, namely, natural forest, forest plantation, upland rice field, mixed crop cultivation area and abandoned area. Then zoning for land used management was planned using the GIS technique to establish a developing project for rural development and watershed conservation. Agroforestry systems and terracing cultivations were used for protection from soil loss, land slide and erosion. Many kinds of multipurpose tree species, fruit trees and nitrogen fixing trees (NFTs) such as Acacia mangium, Broussonetia papyrifera, Bambusa sp., Musa sp., Persea americana, Macadamia integrifolia, Magifera indica, Litchi chinensis, Lansium domesticum, etc. were introduced for ecological reasons and economical profits.

No.P-12

Serious Urban and Forest Landscape Destruction by the Tropical American invasive Weeds in India

Ravinder K. Kohli, H. P. Singh and D. R. Batish

Centre for Environment, Department of Botany, Panjab University, Chandigarh 160014, India

Key Words: Ageratum conyzoides, allelopathy, Parthenium hysterophorus, Lantana camara, invasive weeds, tropical American

The fast rate of development has led to shortening of the distances not only for humans but also for other species. As a consequence, aggressive aliens species have adversely influenced the community structure and its utility in any area they invade. Apart from disturbing the natives and shattering the ecosystem integrity, they affect the economy and the stability and beauty of the landscape. In India, a number of such invasive exotic weeds especially those from tropical America viz. Parthenium hysterophorus, Lantana camara and Ageratum conyzoides, are troublesome and have caused adverse ecological, economic and social impact. These noxious weeds though can be seen growing in different landscapes but are luxuriantly localized in urban, forest and cultivated areas, respectively. P. hysterophorus (Asteraceae) commonly known as congress grass is perhaps the most troublesome for urban and rural India. Being a rapid colonizer it replaces native vegetation fast apart from causing a number of health problems such as skin allergy, rhinitis and irritation to eyes of the residents in the vicinity. Likewise, it causes fodder scarcity apart from being unpalatable and toxic to livestock. Lantana camara (Verbenaceae), another serious and worst tropical American weed has encroached a large area of forested landscape and virtually replaced the floor vegetation, declined tree growth and spoiled the beauty and economy. Further because of its bushy and spreading type of growth it obstructs forest operations. Third weed, Ageratum conyzoides (Asteraceae) has infected the agricultural fields. It is also a serious invader of rangeland also where it competes with native grasses causing scarcity of fodder. In view of the enormity of the problem, managing any of these hazardous weeds is really difficult

It is proposed to discuss the reasons of the problems, share the control tactics adopted and the partial success achieved

No.P-13

Central European natural forest bird communities and guilds: Case study

in West Carpathians fragmented forest landscape

Kropil Rudolf

Department of Forest Protection and Wildlife Management, Faculty of Forestry, Technical University, Masarykova 20, SK - 960 53 Zvolen, Slovakia

Key Words: communities, biodiversity, forest reserves, mapping technique

Primeval and natural Central European forest remnants of the West Carpathians, conserved till the present time as reserves, represent bio-centres in fragmented forest landscape and simultaneously a unique opportunity to study original communities or natural principles in general. I analysed the structure of bird communities in different Central European natural and primeval forest reserves in the altitudinal range from the lowland floodplain forest, through oak, beech and mixed mountainous (represented by the fir-beech and fir-spruce-beech stands) to the spruce forests (altogether 77 census/years within 25 plots of the high timber productivity). Bird censuses were carried out from 1990-2002 using the combined mapping technique. Different aspects of community structure such as species richness, overall density and diversity as well as guild organisation are discussed in detail. The overall density declines with an increasing altitude from 151.7¡À22.8 breeding pairs/10 ha in the floodplain forests to 42.2¡À10.5 in the spruce forests, where only differences between the beech and mixed mountainous forests were not significant (ANOVA, Duncan test, p>0.05). Although the highest values of species richness and diversity were found in the oak primeval forests (34.7¡À2.9), only differences between the spruce (20.4¡À7.9) and other forests were statistically significant (ANOVA, Duncan test, p<0.001). More than 80 % of birds fed on invertebrates and almost half of them were foliage insectivores. High densities of hole and ground nesters are characteristic for primeval and natural forests under study in comparison with man-made forests and non-forest habitats. The results underlined that the primeval and natural forest reserves are of high importance as bio-centres and bio-corridors in fragmented forest landscape of the West Carpathians.

No.P-14

Forest fragmentation due to long-term forest-limit retreat

in northern Finland - the roles of climate and man

Samuli Helama

Department of Geology, Division of Geology and Palaeontology, P. O. Box 64, FIN-00014 University of Helsinki, FINLAND

Key Words: Forest-limit, Scots pine, palaeobotany, climate change

The coniferous forest-limit in Finnish Lapland is often broad, transitional zone from forest to isolated trees, being largely a polar forest-limit. The long-term forest-limit retreat of Scots pine (Pinus sylvestris L.) has caused the fragmentation of remaining forest stands. Potential reasons of this retreat include climatic and anthropogenic factors as well as their combination. In this paper these factors are presented together and their impacts are reviewed according to the results of separate ongoing and completed studies.

Numerous subfossil pine trunks have been collected from the bottom sediment of small lakes in the region (Eronen et al. 2002). Abundant presence of the subfossils dates back to 1000 years ago, thereafter the record implies the long-term thinning of the forest-limit region. In addition, big trunks of subfossil pines are found from the localities where pine is not present today (Eronen et al. 2002; Helama et al. 2004). Local temperature reconstruction, derived from the tree-rings of the subfossil pines, shows long-term temperature decrease, pointing to the general deterioration of the climate in the forest-limit region. Regeneration of pine at forest-limit is significantly depending on the summer temperatures. Therefore, it is evident that described decline of temperatures has influenced negatively the regeneration success of pine in the region. It is notable that temperatures during the entire 19th century were too low for major pulses of pine regeneration. Rise of the temperatures during the 20th century, however, has caused the amelioration of the forest-limit pine success.

In northern Finland, the increased settlement spreading from the south and correspondingly increased forest use started in the 18th century and continues onwards (Veijola 1998). Anthropogenic impacts on the forest-limit include mainly felling, fire, grazing and erosion (Mattsson 1995; Veijola 1998). Anthropogenic influence commonly occurs over restricted areas but in the aggregate the factors and their influences may bear broader implications. According to Mattsson (1995), the forest-limit has retreated due to various anthropogenic activities roughly 30 km in the region during the past 200 years.

As a conclusion, it is evident that in particular the long-term retreat of forest-limit, and the corresponding fragmentation of the remaining valley forests, have both occurred due to climatic deterioration. This in turn has caused even centennial gaps in the pine regeneration in the region. It is likely that the joint-occurrence of such a gap during the severe 19th century with beginning of increased anthropogenic use of the forests, considerably intensified the negative impact of each other. The study is not aiming to neglect the influence of man but points to the importance of climatic fluctuations as an agent strengthening anthropogenic environmental changes. Long-term perspective can be recovered for example by means of palaeobotany.

No.P-15

Seed production of Fagus crenata in relation to patch size and geographical distribution of cool-temperate deciduous forests

in Shikoku Island, southwest Japan.

Shigeo Kuramoto¹, Yoshiyuki Inagaki¹, Atsushi Sakai¹, Eiji Kodani¹, Katsuya Masubuchi², Shigeho Sato¹, Osamu Kobayashi³ and Yasumasa Hirata¹

1. Shikoku Research Center, Forestry and Forest Products Research Institute

2. Laboratory of plant ecology, Faculty of science, Kochi university
3. Experiment forest, Faculty of agriculture, Ehime university

Key Words: Beech, deciduous forest, Fagus crenata, Forest fragmentation, Seed production

Seed production of Fagus crenata was quantified in Shikoku Island, southwest Japan, in relation to patch size and geographical distribution of interspersed cool-temperate deciduous forests. Seed fall was measured for the 165 beech trees in 22 forest sites (8 trees per forest site in average) from 2002 to 2003. Patch size and geographical distribution of interspersed cool-temperate deciduous forests were analyzed, based on the natural environment GIS and land mesh climate data. The amount of seed production was substantially different among 22 forest sites and between years. Generally, mast-crop was observed in 2002 followed by poor-crop in 2003. Based on annual seed production of average tree in each sites, pattern of seed production in 22 sites was clarified into three groups; regular mast crop site, non-synchronized site (mast crop was observed in 2003) sites and poor-crop site. Regular mast-crop sites were distributed in central mountains of island, where large forest patches were distributed. Non-synchronized and poor-crop sites were distributed in peripheral mountains of central mountains, where small forest patches were interspersed. We also found the log-linear relationship between patch size of cool-temperate deciduous forest and magnitude of seed production (defined as average seed production in good crop year for each sites) across 22 forest sites. Patch size of cool-temperate deciduous forest was also log-linearly correlated with altitude across 22 forest sites. Central mountains of Shikoku Island were higher than its peripheral mountains. Therefore, patch size of cool-temperate forests was potentially large in central mountains. In lower mountains, fragmentation of cool-temperate deciduous forest caused by conifer plantation also makes patch size smaller besides potential limitation by geomorphologic condition. Patch size of cool-temperate deciduous forest, which was potentially limited by altitude of mountains affect magnitude of seed production of beech trees.

No.P-16

Changes in insect species diversities along a chronosequence of secondary broad-leaved forests in temperate Japan

Shunichi. Makino¹, H. Goto², T. Inoue¹, M. Sueyoshi⁴, K. Okabe¹, M. Hasegawa³,

K. Hamaguchi¹, H. Tanaka, and I. Okochi¹

1.Forestry and Forest Products Research Institute, Matsunosato 1, Tsukuba, 305-8687, Japan

2.Kyushu Research Center, Forestry and Forest Products Research Institute, Kumamoto, 860-0862, Japan

3.Kiso experimental station, Forestry and Forest Products Research Institute, Nagano 397-0001, Japan

4.Department of Systematic Biology, Entomology Section, PO Box 37012, National Museum of Natural History, Smithsonian Institution, Washington, DC 20013-7012 USA.

Key Words: secondary deciduous forests, arthropods, plants, chronosequence, species richness

A biodiversity monitoring program has been conducted in the Ogawa Forest Reserve and its vicinity situated in a region in temperate Japan. Forests of the region are characterized by a mosaic of secondary deciduous stands of various ages scattered among plantations of conifers, Japanese cedar and hinoki cypress. The monitoring results showed three different types of change in species richness in response to the stand age. In Type I (butterflies, tube-renting bees and wasps, hoverflies, fruit flies, and logicorn beetles), the species diversity was largest in open area just after clear cutting, decreasing with the stand age; in Type II (mites on mushrooms), more species were collected in older stands as compared with younger ones; and in Type III (moths, oribatid mites, collembolas, carabid beetles, and ants), the number of species did not change greatly with the stand age, though species compositions varied as revealed with the standard ordination analysis. These indicate that combinations of stands of different ages, or heterogeneously arranged stands, contribute to maintenance of general insect biodiversities at the landscape level.

No.P-17

Qualitative spatial representation of forest landscape with semantic modeling of topography using digital elevation models

Toshiya Matsuura, Masamu Aniya, Makoto Yokohari, and Wajiro Suzuki

* Graduate school of life and environmental sciences, University of Tsukuba, Japan,

** Graduate school of systems and information engineering, University of Tsukuba, Japan.

*** Forestry and forest product research institute, Japan.

Key Words: ualitative spatial representation, Semantic modeling, Topography, Digital elevation models, DEM, Forest landscape

Topography is not only the basic factor which influences forest landscape but also forms the basis for its recognition. Various indices such as local slope, hydrological and topographic features have been developed using digital elevation models (DEM) for the applications in hydrology, forestry, and landscape ecology. However, there are several differences between the existing topographic indices using DEMs and the recognition of topographic features from map interpretation and fieldwork. In this study, we developed algorithms for representing semantics (meaning of words) of topography for describing topographically different regions. A 10 m DEM was used in this study. Algorithms were developed using programming language, Visual Basic 6.0, which are composed of three steps: 1) delineation of critical boundaries on surface such as the top and the bottom of slopes; 2) connecting the adjacent critical boundaries of slopes; and 3) querying the topographic contexts based on mereotopology (mereology, the theory of part-whole relations: and topology, the theory of adjacency). Based on the developed algorithms, several topographic features such as valley floors, ridge lines, terraces and uplands, and the upper and lower slopes of the valley sides were described for test sites selected from areas of the lowlands, dissected uplands, river terraces, hills, and mountains in the Kanto district, central Japan. Then we applied these algorithms to the knowledge representation of forest landscape. Some of the qualitative spatial representations of forest landscape, such as the riparian vegetations on the foot of the slopes adjacent to the narrow sloping valley floors were described. These algorithms can be one of the bases of knowledge representation of forest landscape in complex topographic regions for the purposes of their conservation and restoration.

No.P-18

The Environmental Selectivity to the Land Use and the Lower-Layer Vegetative Structure by Raccoon dogs, Nyctereutes procyonoides viverinus, in urban area

Youichi Sonoda and Noboru Kuramoto

Graduate school of Agriculture, Meiji University, Japan

Key Words: Raccoon dogs, GIS, Radiotelemetry method, Environmental Selectivity, Land use, Lower-Layer Vegetative Structure

We analyzed the relationship between changes in the distribution of Raccoon dogs, Nyctereutes procyonoides viverinus Temminck, and habitat preferences using GIS. Applying principal component analysis and cluster analysis, we classified land use in Kanagawa Prefecture in 1984 and 1994 into 8 types: woodland, farmland, residential, urban, industrial, mosaic of woodland and farmland (MA-I), mosaic of woodland, farmland, and residential (MA-II), and mosaic of farmland and residential area (MA-III). We used Ivlev$B!G(Bs (1955) selective coefficient to understand the relationship between distribution and landscape classification. Raccoon dogs significantly preferred farmland, MA-I and MA-II in 1982, 1987, and 1992. The distribution of Raccoon dogs is expanding into urban land (residential, urban, and industrial areas) from $B!H(Bnatural$B!I(B areas. This shows that Raccoon dogs can use urban land as habitat in addition to $B!H(Bnatural$B!I(B areas. At the next step, We studied the impact of the lower-layer vegetative structure on habitat selection by raccoon dogs. First, we calculated$B!H(Ba plant biomass index$B!I(Bbased on the area, sociability, and height of each vegetation unit of the lower layer, and determined the relationship between the distribution of lower-layer vegetation and the geographical features of the site. Next, we used radiotelemetry to evaluate environmental selectivity by raccoon dogs. We related the structure of the lower-layer vegetation to the habitat selected by the raccoon dogs. Our analysis revealed that the raccoon dogs had a significant preference for inclined ground with a high plant biomass index and a high occurrence of Pleioblatus chino. Thus, the structure of the lower-layer vegetation in a forest is as significant as the forest type in restricting the habitat of raccoon dogs. Therefore, if raccoon dogs were designated as for conservation, it would be necessary to preserve large areas of Pleioblatus chino with high plant biomass in open spaces.