Influence of climate, elevation, and land use in regional herpetofaunal distribution in Tochigi Prefecture, Japan

Understanding the relationship between the distribution of organisms and their environment is important for regional conservation planning. While most herpetofaunal community studies address environmental influence on species richness, few studies describe environmental influence on species composition at geographic scales. Field surveys of herpetofauna were conducted during 1994-2000 in Tochigi Prefecture, which covers ca. 6400 km of mainland Japan and includes a wide range of elevation, annual mean temperature, and human population densities. We evaluated whether (i) regional herpetofauna are distributed along a gradient of climate and elevation, as well as land cover and land use, and (ii) ‘climate/elevation’ and ‘landcover’ variables differ in their relative contribution. Ordination indicated a strong geographic gradient in species composition that was correlated with climate, elevation, land cover and land use. Reptiles were strongly influenced by land cover and land use (residential, agricultural, and forested area), while amphibians responded to all gradients (climate, elevation, and land cover and land use). All species with declining populations listed in the Prefectural Red Data Book were associated with mid-elevation regions, where anthropogenic modification is most intensive. All exotic species were strongly associated with anthropogenic gradients. Nomenclature: Goris and Maeda (2004) and Ota (2005). Abbreviations: CA – Correspondence Analysis, CCA – Canonical Correspondence Analysis, pCCA – Partial Canonical Correspondence Analysis. The purpose of this paper was (i) to test whether regional herpetofauna are distributed along climatic gradients (temperature and precipitation) and elevation, as well as land cover and land use (forest cover, residential area, agricultural area, area of lake/pond/marsh), and (ii) to determine the relative importance of ‘climate/elevation’ and ‘landcover’ variables in explaining herpetofaunal distribution. We also evaluated whether amphibians and reptiles respond differently to the environment due to their basic physiological differences. Finally, we assessed the relationship of particular species with human activities (e.g., deforestation and habitat conversion) and the distribution of herpetofauna species of conservation concern. Materials and methods Study area and species dataset Tochigi Prefecture (36°34’ N and 139°53’ E) covers ca. 6400 km of Honshu, the main-island of Japan. Despite its relatively small size, the prefecture exhibits diverse landscape features and striking environmental gradients. The northwest region of this prefecture consists mainly of high mountains (up to 2578 m in elevation) and hilly terrain (600-1000 m) occurs in the eastern region. Plains and low hills (14-200 m) dominate the topography in the central and southern regions. Tochigi belongs to the humid temperate zone. Annual average temperatures are 12.5 C in central and southern regions and 7.5 C in northern region (Tochigi Prefecture 2001). Annual average precipitation is 1200 mm in the central and southern areas and 2000 mm in the northern area (Tochigi Prefecture 2001). Twelve trained investigators (including one of us, T.H.) conducted surveys during 1994-2000 throughout the prefecture. The surveys were originally aimed to gather basic data for conservation of prefectural herpetofauna, and subsequent production of their summary volume for Tochigi Prefecture (Tochigi Prefecture 2001). The investigators covered all potential habitat types, including forests, grassland, wetland, river/stream, rice paddies, crop field, residential areas, and commercial districts. Searches primarily involved flipping rocks, logs, and debris, dip-netting in water, and frog and toad calling survey. Surveys were conducted both day and night time to find all potentially-occurring species. The resulting data set consists of 15 reptile and 18 amphibian species (Appendix 1) from the total of 2489 localities (amphibians = 1960, reptiles = 529). A summary of this survey (without statistical analyses), including additional data obtained from questionnaires, published accounts, and museum records, is published in the governmental report volume (Tochigi Prefecture 2001). Environmental variables We obtained data for 9 environmental variables from governmental sources (available from Municipal Governments and/or online from their official homepages), that were of potential relevance for herpetofaunal distribution and other herpetofaunal studies (e.g., Raxworthy et al. 2003). All the variables are based on the means of 19942000 for each municipality. To account for the area variation among municipalities, we calculated each ‘landcover’ variable as a percentage of the total land area of a municipality. We categorized the variables as either ‘climate/elevation’ or ‘landcover.’ ‘Climate/elevation’ included mean annual precipitation, mean annual temperature, mean annual temperature range, and mean elevation, and ‘landcover’ included forested area, cropland, rice paddy field, lake/pond/marsh, and residential area.