Published papers

We studied relationships between bird communities and vegetation structure on isolated woodlots in Tokorozawa City, Saitama Prefecture, Central Japan. Twenty-four sites were located on two isolated woodlots of which ones were 14.5 ha and 7.8 ha, respectively. On wintering season in 1994 bird communities were censused for 20 minutes by point counts method in each site. In each site two 40m long transects and one 50×2m quadrat were established for vegetation survey. Vegetation structure and composition were recorded at 40 points (spaced at 2m intervals) along these lines and basal area of woody plant species (>1.5m) was recorded in the quadrat. Our results showed that vegetation cover density was found to correlate with the bird species richness on wintering season. Based on the species composition of bird, 24 sites were classified into four types namely, type A to D. On type B and D Bush Warbler Cettia diphone and Black-faced Bunting Emberiza spodocephala were characteristically recorded. Vegetation cover density in shrub and herb layers on type B and D sites was significantly higher than that on type A and C sites. Type A had poorest species richness in all types. Vegetation cover density on type A sites was significantly lower than that on type C sites. On type B sites Japanese Green Woodpecker Picas awokera was characteristically recorded. The total basal area of dead trees on type B sites was larger than that on type D sites, however the difference was not significant.

Bird communities on the 20 fragmented woodlots during the wintering and breeding seasons to analyze the factors that influenced the distribution of birds were surveyed in the Tokorozawa City, Saitama Prefecture, Central Japan. Our the results showed that woodlot area is the best predictor of the number of individuals and the species richness. Based on the species composition of bird, the fragmented woodlots were classified into four types: those are core woodlots (type A), woodlots surrounding core ones (type B), woodlots in the field area (type C), and isolated or small woodlots (type D). Besides the woodlot area, the connectivity between woodlots and the landuse surrounding woodlots were recognized as factors influencing avian communities. The necessity of controlling these factors for avian conservation was pointed out.

Multivariate analysis of ecological data has been applied to environmental analysis and evaluation. Recently, some ecologists reported that classification and regression trees (CART) are ideally suited for the analysis of complex ecological data. We use classification tree to analyze the relationship between avian species composition and habitat conditions from 37 study plots located in a riparian area of the Tama-gawa River, Tokyo. The data were comprised of census data of birds and vegetation structural information. First, the study plots were classified by TWINSPAN based on the avian species composition. Then, we tried to recover the grouping of the study plots by classification tree or canonical discriminant analysis using the vegetation structural information so that we could find the relationship between avian species composition and vegetation structure. Classification tree analysis performed almost as well as canonical discriminant analysis. Classification tree models explain variation of a single response variable (here, avian fauna type) by repeatedly splitting the study plots into more homogeneous groups, using combinations of explanatory variables (here, vegetation structural parameters). This structure is simple, suitable for dealing with high-order interactions, so that classification tree can give easily interpretable results. Finally, based on the comparison of classification tree and canonical discriminant analysis, we concluded that classification tree was more suitable than discriminant analysis for landscape evaluation and planning.

Dragonfly species were surveyed in 23 small irrigation ponds in Hokudan-cho, Hyogo Prefectre in order to detect environmental factors influencing dragonfly distribution. Thirty-two species and 1066 individuals were observed from May to October 2001. The number of dragonfly species had no significant relation with water body area of irrigation ponds. Twenty-three ponds were classified to three types by TWINSPAN (Two-way Indicator Species Analysis). The result showed that neighboring woodlots and frequency of cutting glass around pond influenced the component of dragonfly species. Cutting glass is one of the important maintenances of irrigation ponds. It was suggested that the disturbance by traditional farm works such as cutting glass plays an important role to maintain dragonfly species richness on this area.

There is a huge number of small irrigation ponds in the agricultural landscape of the north of Awaji Island, Hyogo Prefecture. Recently, managements of many ponds have been abandoned. However, these ponds are used as habitat by many organisms. They have an important role to maintain biodiversity in this area. Dragonfly (Odonata) species were surveyed in 24 irrigation ponds in Hokudan-cho, the north of Awaji Island. Thirty-seven species were catched and/or observed from June to September 2000. The number of dragonfly species had no relation with water body area of irrigation ponds. Twenty-four ponds were classified to five types by TWINSPAN (Two-way Indicator Species Analysis). The classification of TWINSPAN was analyzed by Classification and Regression Trees using explanatory variables about environmental factors of irrigation ponds. The results showed that altitude, neighboring woodlots, water body area and water quality influenced the component of dragonfly species. Especially, it was important for species preferring edge and/or inside of woodlot that over 45 percents of pond surroundings were adjacent to woodlots. It was also suggested that surrounding land uses influenced the distribution of dragonfly species.