Geographical review of Japan series B Vol. 88(2015) No. 1

This study attempts to propose an alternative framework for examining the transformation process of traditional ethnic neighborhoods and understanding the changing spatial structure of ethnic communities in multiethnic cities in contemporary North America. Since the beginning of immigration, the Portuguese have been residing in Toronto for more than half a century. This community now faces a generational change. Toronto’s Portuguese community is examined by focusing on the ethnic functions that comprise its residences, businesses, and organizations in order to dismantle the complicated spatial structure of today’s ethnic community. From the 1970s through the early 1980s, such ethnic functions were concentrated in Little Portugal, located near downtown Toronto. During the 1980s, however, Portuguese residential space began to spread from Little Portugal to Toronto’s northern corridor and western suburbs. After the mid-1990s, Portuguese organizations relocated to the northern corridor. Moreover, the number of Portuguese businesses has decreased since the beginning of the 2000s, consequent to the aging of first-generation Portuguese and the inflow of non-Portuguese people, or gentrifiers. In other words, ethnic functions relocated from Little Portugal in multiple stages. However, Portuguese entrepreneurs still manage approximately half of the businesses in Little Portugal, and this traditional ethnic neighborhood remains the core area of the Portuguese community. Today, the Portuguese in Toronto utilize plural spaces depending on the content of their activities. Although the Portuguese community is spatially dispersed, its social ties are maintained on the basis of ethnicity, and these three spaces are thus closely connected with each other.

Riparian areas are unique habitats that contribute to biodiversity, so they have been focused on in many regions; however, subalpine riparian forests have hardly been examined in Japan. We investigated the micro-landform structure and spatial pattern of tree distributions in a V-shaped valley at 2,000–2,200 m a.s.l. in the Minami Alps, central Japan, using a 0.42 ha core-plot and 16 belt-transects set in a headwater area of Norogawa River. As riparian topographical components, channel, floodplain, scarp of terrace and terrace were detected, which were arranged roughly from lower to higher elevations from streams, as well as mountain slope as a micro-landform unit outside the riparian area. Single-layered floodplain, conspicuous terrace segments and the probable lack of a lower sideslope were identified as features that differ from those found in previous studies on other climatic/large-scale geomorphological conditions. The distribution of deciduous species was biased to lower elevations, with the representatives Salix cardiophylla var. urbaniana on floodplains and Alnus matsumurae on scarps of terraces. These micro-landform units were recognized as a riparian zone in terms of vegetation. Meanwhile, climatic climax evergreen conifers, mainly Abies veitchii, Tsuga diversifolia and A. mariesii, dominated not only mountain slopes but also terraces, indicating that terraces are upland areas in terms of their vegetation. A much smaller area, low species diversity and an assumed direct succession from pioneer to climax phase, because of poor long-lived riparian species capable of forming a mid- or late successional phase, were properties differing from those found in previous studies on other climatic/geomorphological conditions.

This study estimated near-ground air temperature (T_a) from Terra/ASTER-obtained land surface temperature (LST) for different land cover categories in a hilly region under a clear and calm night in winter. Temperature differences (ΔT) between LST and T_a were related to the land cover categories and their spatial extents. In the extended area of forest land, spatially averaged ΔT in the area (ΔT_m) was −0.1°C, and the relationship between LST and T_a showed less variability (standard deviation of ΔT (SD)=0.8°C) and high correlation (coefficient of determination between T_a and LST (R²)=0.70). This result suggests we can evaluate T_a directly from LST in this area. In the extended area of bare land, ΔT_m, SD and R² were −1.5°C, 0.9°C and 0.85, respectively, which indicated that accurate estimation of T_a is possible by correcting LST by ΔT_m. In the extended area of artificial land, ΔT_m showed statistically non-significant trend. In the case of intermingled area of more than one type of land cover, ΔT_m, SD and R² were −0.8°C, 0.9°C and 0.71 in the forest land of the intermingled area, which suggested that we can estimate T_a in this area precisely by correcting LST by ΔT_m. In the bare land of the intermingled area, though ΔT_m was −1.4°C, large variations (SD=1.4°C) and low correlations (R²=0.33) were found.