Exploitation of the semi-arid region of Australia began in the late 19th century. Major expansion of farm land started at the beginning of this century. The landscape covered with the dense mallee changed to wheat fields or green pasture lands. The change was very rapid and drastic.
In this paper the authors examine the characteristics of marginal land for agriculture in the semi-arid region by analyzing the relationship between wheat growing and rainfall. The condition of marginal land for agriculture and the reality of wheat growing in the semi-arid region will be clarified based on the analysis of the correlation between wheat yield (bushels per acre) and winter rainfall from June to September, which coincides with the growing season for wheat. The Murray Mallee region of South Australia in the period 1907-1946, was selected as the main study area (Fig. 1). This study uses the following sources: “Statistical Register of South Australia-Production” compiled in the South Australian Parliamentary Paper, and “Report of Monthly and Yearly Rainfall, 1907-1946” collected by the Bureau of Meteorology, Australia. The 33 hundreds and stations examined are: Cadell (Morgan P. O.), Paisley (Notts Well), Waikerie (Waikerie Lands), Holder (Maggea), Moorook (Moorook), Nildottie (Swan Reach), Mantung (Caliph P. O.), Bandon (Copeville) and Mindarie (Mindarie) in County Albert, Murtho (Wilkadine), Paringa (Renmark P. O.), Gordon (Taldra P. O.), Pyap (Pyap), Bookpurnong (Loxton P. O.), Allen (Alawoona), Kekwick (Paruna) and McGorrery (Meribah P. O.) in County Alfred, Vincent (Perponda), Wilson (Borrika), McPherson (Sandalwood), Hooper (Wynarka), Marmon Jabuk (Karoonda), Sherlock (Moorlands), Roby (Coomandook), Peake (Peake), Price (Parrakie P. O.), Livingston (Kiki) and Coneybeer (Coonalpyn) in County Buccleuch, Billiatt (Gurrai P. O.), Kingsford (Karte), Bews (Lameroo P. O.), Parilla (Parilla) and Pinnaroo (Pinnaroo P. O.) in County Chandos (Figs. 2 and 3). The statistics from these hundreds and stations provided sufficient working data.
Originally, wheat yield was not stable here because of the influence of variable rainfall (Fig. 4). In the semi-arid region, if rainfall is plentiful, the wheat yield increases proportionally. But if there is a scanty rainfall, mainly in winter, the wheat growing will suffer deadly damage. The characteristics of the correlation between annual wheat yield and annual winter rainfall (1907-1946) can be classified into four types: Types I-IV (Figs. 7, 8, 9 and 10). Fig. 11 shows the distribution of the four types. The data shows that hundreds included in the same type tend to be in close proximity.
D. W. Goyder, the Surveyor General of South Australia, defined the extent of servere drought in 1865. This is called “Goyder's Line of Rainfall”. It was a base line against which the safe and the unsafe areas for semi-arid agriculture were measured. It has a considerable validity even today. In the Murray Mallee region, “Goyder's Line of Rainfall” was drawn from Swan Reach in the hundred of Nildottie to the south of Peebinga in the hundred of Peebinga near the Victorian boundary.
From analysis of the four types shown in Figs. 7, 8, 9 and 10, it is clear that “Goyder's Line of Rainfall” is not a fixed or static line but a boundary belt which moves dynamically within the Type II area, that is the safe and the unsafe wheat growing area. Also it can be suggested that the wheat growing conditions in the marginal land change drastically at the border, the difference being as high as 100 to 120mm of winter rainfall from June to September.

抄録全体を表示

Citation analysis is a bibliometric method using reference citations found in scientific papers as the primary analytical tool. This paper attempts to clarify the communication network of one field in geography by means of citation analysis. Central place studies, originating from Christaller and Lösch, are selected for this purpose.
Citation is taken to represent four indicators as follows: 1) quality, 2) recognition, 3) diffusion and communication network, and 4) utilization. First, this paper reviews various measures used in citation analysis, such as bibliographic coupling, link connection, and co-citation, which have developed in the fields of Sociology of Science, and Library and Information Science. Multivariate data analyses, especially cluster analysis, factor analysis and multidimensional scaling (MDS), make it possible to deal with a large number of citation data such as the Scientific Citation Index (SCI) and the Social Sciences Citation Index (SSCI).
Second, citation analysis used in the field of geography is reviewed (see Table 1). According to the indicators mentioned above, the studies are divided into three categories: 1) quality of research, productivity and ranking, 2) professional recognition and reward, and 3) communication network and diffusion of ideas.
Third, to examine the communication network of geographers, the early central place studies are considered. In the present study, the citation frequency of 98 papers by 57 geographers which are taken from “Central Place Studies” (Berry and Pred, 1965) are counted directly, and an asymmetric matrix (57×57) is constructed. Citation frequencies are transformed into “distance”; hence the distance between geographers is analyzed by MDS, which assigns relative locations in two-dimensional space: the scaled configuration of highly cited and co-cited geographers would be a concrete representation of the communication network. It is expected that the 57 geographers comprise subgroups based on nationality, specialty and/or school.
In the present study, seven clusters are identified: 1) the Lund school, 2) the Chicago school, 3) the Washington school and researchers with theoretical and quantitative approaches, 4) English researchers, 5) CBD researchers, 6) researchers concentrating on the internal business structure of the city, and 7) marketing researchers (see Fig. 4).
As a result, reference space fundamentally consists of two axes: “quantitative analysis vs. qualitative description” and “European geographers vs. American geographers”. When properly used, citation analysis can introduce a useful and effective measure of objectivity into the classification and evaluation of science and recover a communication network among researchers.

抄録全体を表示