Hiroaki Ishida

In order to clarify the effects of forest management on the composition and richness of all vascular plant species in lucidophyllous (evergreen broad-leaved) forests, secondary lucidophyllous forest stands regenerating after clearcutting (ca. 40 years old), which were previously clear-cut logged in 15-20 year cycles, were compared with primary lucidophyllous forest stands in two altitudinal zones (Castanopsis zone, ca. < 350 m a.s.l.; and Quercus zone, ca. > 350 m a.s.l.) on Tsushima Island, Japan. For each altitudinal zone, the occurrence frequency of each species in the primary and secondary stands was compared using Fisher's exact test. Eighty-one species were shown to have a significant difference, accounting for about 49% of all species examined (n = 165). In both zones, many species found in the primary stands were absent or rare in the secondary stands. We classified all species into one of two groups: component species of lucidophyllous forests (lucidophyllous elements) and other species (non-lucidophyllous elements). The secondary stands had lower total species richness (the number of all species per 100 m(2)) than the primary stands, regardless of the altitudinal zone. The lower total species richness in the secondary Castanopsis zone stands reflected a decrease in the number of lucidophyllous elements, whereas the lower total species richness in the secondary Quercus zone stands primarily reflected a decrease in the number of non-lucidophyllous elements. The beta diversity between the altitudinal stand types, measured as Whittaker's index beta(w) was lower in the secondary stands than in the primary stands. This difference was primarily due to a lack of species characteristic of each altitudinal zone. (c) 2005 Elsevier B.V. All rights reserved.

In order to clarify that Ligustrum lucidum has invaded the wild in Japan, we researched by using literatures on how so many Ligustrum lucidum had been imported into Japan as a roadside tree over the last 23 years, and how many out of the 123 major rivers in Japan Ligustrum lucidum was found in. Then we investigated existence, number of individuals, distribution density and the growing condition of Ligustrum lucidum at the natural and artificial green districts in two new residential areas in western Japan. In result, we revealed that about 15 million Ligustrum lucidum were imported to Japan for the last 23 years and that Ligustrum lucidum has been distributed at 25.2 percent of the major rivers in Japan. 31.9 percent of natural green districts have been invaded by a lot of individuals of Ligustrum lucidum. The density of these was very high, and Ligustrum lucidum has grown well in natural and artificial forests. To our knowledge, we conclude that Ligustrum lucidum is an invasive alien plant in Japan.

Other than artificial forests, three major types of lucidophyllous (evergreen broad-leaved) forests are found in Japan, according to the degree of human impact: (1) primeval; (2) natural-those that have been subjected to occasional low-impact human disturbance, such as tree or understory cutting, and to a high degree of fragmentation by human development; and (3) secondary-coppices that was used as a source of firewood and charcoal. In this study, species composition and richness were compared among primeval, natural, and secondary lucidophyllous forests on Tsushima Island, Japan, to clarify the effects of human disturbance and forest fragmentation. The vegetation of 76 plots of 100m^2 (30 primeval, 20 natural, and 26 secondary plots) was investigated. The results of detrended correspondence analysis showed that: (1) the species composition clearly differed among the forest types; and (2) the species composition of the natural forests was in between the compositions of the primeval and secondary forests, showing similarities to both. The differences in species composition among the three forest types seemed to greatly reflect the degrees of both past human disturbance and forest fragmentation. Distylium racemosum, which has a low coppicing ability, dominated in the primeval forests, but this was not or rarely the case in the natural and secondary forests. Species richness (number of species per 100m^2) differed significantly among the forest types, from highest (primeval) to lowest (secondary). The number of all species per 100m^2 did not have a clear relationship with forest fragmentation. However, the number of species whose distributions were biased toward the primeval and natural forests was highly and positively correlated with patch area of the primeval and natural forests, suggesting that forest fragmentation is partly responsible for the differences in species richness among the forest types.

In order to clarify the effects of human disturbance and forest fragmentation on the lucidophyllous forest, we compared species composition and richness among the following three types of lucidophyllous forest in Miyazaki Prefecture in southeastern Kyushu, Japan: (1) primeval lucidophyllous forest; (2) natural lucidophyllous forest, i.e., small patches of shrine forest in which no periodic clear cutting has occurred but where there has been occasional, low-impact human disturbance as well as a high degree of fragmentation by human development; and (3) secondary lucidophyllous forest, i.e., coppices that were used as a source of firewood and charcoal. We set a total of sixty-two 100m^2 plots (11 primeval plots, 19 natural plots, and 32 secondary plots) at the study sites. In each plot, all vascular plant species were identified. The species composition clearly differed among the three forest types; that of the natural forests was intermediate between those of the primeval and secondary forests. Species composition was strongly influenced by both human disturbance and forest fragmentation. Compared to the primeval forests, species that prefer shade and mesic soil and epiphyte species were absent or rare in the natural forests. The number of species per plot for all species and that for the component species of lucidophyllous forests tended to be lowest in the primeval forests, followed in order by the natural and secondary forests. The number of species that occurred disproportionately in the primeval and natural forests when compared to the secondary forests was highly and positively correlated with forest patch area for these two types of forest, indicating that forest fragmentation causes the decrease in species richness.

The coppice forests (summer-green secondary forests) in Sanda city, Hyogo Prefecture were managed by cutting lucidophyllous shrubs and bamboo grasses for the purpose of creating the physiognomy of summer-green high forests, increasing the species richness in the forests and proposing a method of the management of the coppice forests by citizen participation. Six quadrates of 100m^2 were set in the forests, which were identified with Quercetum variabili-serratae. Persons ranging from 8 to 15 investigated the floristic composition and tree census, and managed the vegetation of each quadrate. In this method of forest management, lucidophyllous shrubs and bamboo grasses were cut one species after another species, and the time required for cutting and the number of individuals each species were recorded. The basic rate needed for the above mentioned management work of the coppice forest by citizen participation was 5.1m^2 per man-hour.

The lucidophyllous forests at Shiratori, Kurino and Tatera in Kyushu, Japan, were investigated in order to clarify differences in species composition and species richness between closed canopy stands (mature phase) and canopy gap stands (gap phase). The mature phase was mainly characterized by the presence of many epiphytes which were component species of lucidophyllous forest (elements of Camellietea japonicae). The gap phase was primarily characterized by the presence of summergreen trees, summergreen shrubs, summergreen climbers and others which composed Rosetea multiflorae, Fico-Mallotetalia, Fagetea crenatae and Artemisietea principis. The gap phase was differentiated not only by the above species, but also by the component species of lucidophyllous forest. In the gap phase, the species richness of all occurring species and that of component species, except for epiphytes of lucidophyllous forest, were higher than those in the mature phase. Canopy gaps play an important role in the maintenance of species richness of component species of lucidophyllous forest as well as the process of lucidophyllous forest regeneration.

In order to find out the methods of increasing species richness of lucidophyllous-artificial forest, we investigated that forest in Nanko power station on littoral district, Osaka city. As the result of comparison of life form spectrum between lucidophyllous-artificial forest and lucidophyllous-natural forest, it became clear that lucidophyllous-artificial forest lacks in dwarf shrub, perenial herb, climber, epiphyte and others. The 693 plants of 44 species belonging to those life forms were planted on three testing sites as a seed sources in plant species flow for species richness in the forest from November 1999 to June 2002. We investigated growth condition of those plants. By the reason of little precipitation in summer in 2000, 2001 and 2002 or winter in 1999, 48% of the plants withered. The withering ratio of Cymbidium goeringii, Ardisia japonica, Ardisia crenata, Liriope platyphylla and others were low. It seemed that planting of dwarf shrubs and perenial herbs which had tolerance to aridity was effective, as a method of increasing species richness of lucidophyllous-artificial forest.

A lucidophyllous forest in Kawanaka, Aya, Miyazaki Prefecture, was studied in order to clarify the distribution of the lucidophyllous elements and species richness according to micro-scale geomorphic and the physical-soil conditions. In this area, 5 micro-scale geomorphic units (foot slope, lower sideslope, upper sideslope, crest slope and rocky site) were recognized, and 26 stands for a vegetation survey and 8 plots for a soil survey were investigated. The 26 stands were classified into 5 stand groups according to their microscale geomorphic units. The lucidophyllous elements were assorted into the 8 species groups according to the distributional center of frequency (%) of occurrence and mean coverage (%) of each species along the micro-scale geomorphic gradient. The distributional patterns of the lucidophyllous elements in this area along the micro-scale geomorphic gradient corresponded to those in other areas. It was recognized that in many cases species of the same genus had different distributional centers along the micro-scale geomorphic gradient. Species richness (mean number of species per stand) in each stand group for the micro-scale geomorphic unit ranged from 62.6 to 44.0, and foot slope had the highest species richness which depended on the number of herbaceous plants and lucidophyllous shrubs. On foot slopes, this high species richness seemed to be due to high values of minimum air capacity, indicating highly developed soil conditions.

A check list of flora and endangered species of lucidophyllous forest was made according to red data books and others, in order to clear characteristics of the flora and critical conditions of the endangered species of lucidophyllous forest at Kyushu and northward in Japan. The number of component species of lucidophyllous forest in each prefecture had a high positive correlation with low temperature during winter. The life form compositions of lucidophyllous elements were studied. The life form of the lucidophyllous elements was classified into 19 categories. The value of lucidophyllous tree was about 24 per cent. The dominant life form categories were forest floor herb and epiphytes. Species richness of lucidophyllous forest depended on many ferns and orchids in forest floor herb and epiphytes. About 30 per cent of the total number of component species of lucidophyllous elements is endangered. The endangered species were, for the most part, ferns and orchids.

Nineteen fragmented secondary forests distributed in the urban landscape of the Senri Hills in Osaka Pref. were investigated. The relationship between the number of secondary species (species showing habitat preference for secondary forest) and area of the forest was analysed using the species-log area and the log species-log area models, and in both models the correlation between the number of species and area was related by a highly significant linear regression. Compared with the fragmented secondary forests in Flower Town, Hyogo Pref., the forests in the study area were found to have fewer species. It seemed that the major causes for this were the differences between the two areas in the time and degree of previous use for secondary forest, or in the time of isolation. The secondary species were classified into three groups by the similarity of distributional pattern with respect to the decrease of area. Herbaceous species that prefered mesic habitat tended to be absent in smaller forests, probably because of the dryness of the soil and the lack of mesic microlandforms. In Flower Town, Hyogo Pref. Prunus verecunda, Styrax japonicus, Pourthiaea villosa, Rhododendron reticulatum, etc. were unaffected by the decrease of area, but in the study area they showed a tendency to be absent with decreasing area.

In order to clear the role of source in plant species flow for species richness, we investigated seven lucidophyllous-artificial forests on the littoral districts in the Kinki area. The artificial forests were divided into two types. One was the isolated forest on reclaimed land surrounded by the sea and another was the no-isolated forest encircled with secondary forest. The mean number of occurring species per 100 m²in the isolated forest and the no-isolated forest were 21 and 34 respectively. Species richness in the no-isolated forest was very high compared with the isolated forest. High species richness in the no-isolated forest was due to existence of the adjacent secondary forest which become the source in plant species flow.

Fragmented lucidophyllous forests preserved in the precincts of shrines and temples were investigated to clarify the influence of forest area decrease on species diversity and composition in central Miyazaki Prefecture. The number of component species of lucidophyllous forest had a high positive correlation with the logarithm of the area. The log of the number of species also showed a high positive correlation with the log area. The log area was positively correlated with the number of micro-landform units and the altitude range of the forests. These results suggest that species richness of fragmented lucidophyllous forest is supported by the habitat diversity of the forest. The component species were classified into four groups according to the similarity of their distributional patterns to the area decrease. Forest floor ferns and epiphytes tended to be absent in smaller forests. There were significant positive correlations between the indices of species diversity, Fisher's α and Morisita's β, in the stand (225 m²) and the log areas.

Butterfly assemblages were investigated using route census surveys, at Flowertown in Sanda from April to October, 1994. In order to evaluate the naturalness of vegetation through the species composition of butterflies, we proposed a butterfly index, which indicated the degree of naturalness of the habitat. The values of species diversity, averages of the butterfly index, and values calculated by the Bray-Curtis ordination method were calculated for eight vegetational conditions ; outside fragmented forest ; inside fragmented forest ; outside grassland ; inside grassland ; park ; detached house ; multiple dwelling house ; bare area. The results showed that the species diversity and butterfly index were highest in fragmented forests, and were lowest in the bare area. Ecological distribution of butterflies was classified into four types ; forest-type ; grassland-type ; park-type ; others. Many butterflies belonged to the forest-type classification, since there were many plant communities in the fragmented forests, such as Rhododendro macrosepali-Pinetum densiflorae, Quercetum variavili-serratae, Alnus japonica-Ligustrum obtusifolium Comm. Apanantho-Celtidetum japonicae, Querces acutissima Comm. etc. It seemes that fragmented forests are a very important habitat for butterflies, in spite of the small size of such forests. Butterfly assemblages at Flowertown were compared with other localities, including Mino Park, Hattori-ryokuchi Park, Osaka Castle Park, etc. Mino Park, Mt. Nijosan and Mt. Sanageyama were estimated to be at the secondary vegetational stage, while Osaka Castle Park, Nagai Park and Daisen Park were estimated to be at the urban stage. Although Flowertown was at the secondary vegetational stage in terms of butterfly fauna, it was at the residential-urban stage in terms of butterfly assemblages.