Searching literature started with a basic keyword based (web)search. On regular occasions, online literature resources were consulted for journal entries concerning the topics of the study. Among the online databases were the ISI Web of Science, Elsevier’s ScienceDirect portal, Google Scholar, academic library websites, and publisher specific databases (e.g. Blackwell Synergy, Kluwer Academic).
The most frequently used keywords were “semi urban”, “peri urban”, “sprawl” and “fringe”, in a list of more than 25 keywords. This keyword list was iteratively generated from the literature read: many publications containing the keyword “sprawl”, also contained the keywords “sealed surface” and thus these latter were taken up in the list. From a certain point in reading, the rate to which new keywords entered the list dropped, from this moment on the assumption that the list could be seen as complete was made.
Many essential publications generated a new list of important references: this cited reference based search thus resulted in a list of almost 200 relevant publications. On several occasions, other publications of the same author, different than those listed in references of already selected relevant publications, were also consulted as they frequently cover the same topics.
During the literature study, specific attention was given to topics covering (1) former research or the research base of the study, (2) definitions, concepts, theories and nomenclatures and (3) study methods, especially sampling techniques. The discussion in this paper is primarily based on the findings from (2) and partly from those in (1).
Interesting research programs on urban sprawl, semi-urban areas and the rural-urban fringe are conducted in landscape research across the world, though the biggest, interdisciplinary and international research projects on sprawl and semi-urban areas can be found in Europe, as research programs have been sponsored by the European Union. Certainly worth mentioning in this context is the MOLAND (Monitoring Land Use/Cover Dynamics) research program led by the Institute for Environment and Sustainability – Land Management and Natural Hazards Unit of the Directorate General Joint Research Centre, in cooperation with many other EU-based institutions, such as the European Environmental Agency (European Commission – DG JRC and Space Applications Institute) in Copenhagen (EEA, 2006*) and Eurostat in Luxemburg, or the ESPON (European Spatial Planning Observation Network) 1.1.2. project, finished in 2005, investigating urban-rural relationships in Europe (ESPON, 2006). There are many other research programs induced by European Union research frameworks and followed up by academic institutions that are not part of the EU administration. Some examples of already finished projects include the SCATTER (Sprawling Cities And Transport – from Evaluation the Recommendations) research project (Gayda et al., 2005) in the EU 5th Research Framework, finished in 2004, and the RURBAN (Rural Areas under Urban Pressure) project at Wageningen University (Overbeek and Vader, 2003), also EU 5th Framework and finished in 2005.
At the moment, probably the largest running EU-financed research project on sprawl, semi-urban areas and urban fringe land use is the EU 6th Framework-funded PLUREL (Peri-Urban Land Use Relationships) project. With a consortium of research institutions in 14 European countries and China, the project focuses on the challenges of urbanization and the very specific rural-urban regions, with the Leipzig Charter on Sustainable European Cities as background document (Leipzig Charter, 2007). Other interesting research programs are, for example, the “Peri-Urban Interface” program at the Development Planning Unit of the University College London (PUI), or the BUGS (Biodiversity in Urban Gardens) research program at the University of Sheffield. Although the latter does not contain any direct research on the general urban context issues discussed in this paper, the research projects in the BUGS program do investigate several ecological properties in gardens, being a form of urban land use that does not only exist in city centers, but also far outside the city limits. Much research conducted in the BUGS framework is discussed in the “Urban Domestic Gardens” paper series, for example, by Thompson et al. (2003*); Gaston et al. (2005a*); Smith et al. (2005*).
In North America, many studies on sprawl, semi-urban areas and urban fringe dynamics are funded (and thus steered) by government agencies dealing with non-urban land use like the U.S. Department of Agriculture or the U.S.D.A. Forest Service. An interesting research program is the BES project (Baltimore Ecosystem Study) which started off as the Urban-to-Rural Ecology Gradient Project, run by the Institute for Ecosystem Studies in New York. Examples here are the work of McDonnell et al. (1997*) and Pickett et al. (2004*).
A large scale, interdisciplinary research program generating a considerable amount of literature on sprawl and urban fringes is the CAP LTER (Central Arizona – Phoenix Long-Term Ecological Research) framework of the Arizona State University in Phoenix, conducting multiple long-term studies on the environmental effects of an expanding metropolis in a desert ecosystem. Some examples of papers brought forth from this research project contain the work of Luck and Wu (2002*); Baker et al. (2002*); Jenerette et al. (2007*).
During the odyssey through literature, a list of the locations of case studies and examples used in the papers was created. It is self-evident that many of these locations are dictated by the research framework where the study, containing the locations mentioned, fits in. Some of these research frameworks were discussed in the previous Section 2.2.
In Northern America, Phoenix in Arizona (Baker et al., 2002*; Cook, 2002; Gober and Burns, 2002*; Jenerette et al., 2007*; Luck and Wu, 2002*; Musacchio and Wu, 2004*), New York City (Alfsen-Norodom et al., 2004*; Bugliarello, 2004a*; McDonnell et al., 1997*; Plantinga and Miller, 2001*), Madison in Wisconsin (McMillen, 1989*; Weng, 2007*), Chicago in Illinois (Felsenstein, 2002*; McMillen, 1989*), Atlanta in Georgia (Bourne, 1996*; Gillies et al., 2003*; Wolman et al., 2005*) and Baltimore in Maryland (Musacchio and Wu, 2004; Wolman et al., 2005*) are often used cities to study urban sprawl and related subjects. Statewide studies cover, for example, California (Atkinson and Oleson, 1996*; Cablk and Minor, 2003*; Syphard et al., 2007*), Illinois (Deal and Schunk, 2004*; Sullivan et al., 2004), or New Jersey (Hasse and Lathrop, 2003*; Walmsley, 2006*).
Although still not as much investigated on the sprawl, fringe, or semi-urban subject as European and Northern American urban regions, Eastern Asian study regions are becoming more and more important, evidently because of the very rapid urbanization that is happening in some southeastern countries and especially in China. Examples of papers using cities and regions in Asia include Bangkok (Madhavan et al., 2001*; Yokohari et al., 2000*), Bangalore (Sudhira et al., 2004), Shanghai (Zhang et al., 2004*), or Singapore (Yang and Lay, 2004*).
In Europe, frequently used study areas are the Leipzig region in Germany (Böhm, 1998*; Haase and Nuissl, 2007*; Hümmeler, 1998*), Leeds (Freeman, 1999; Perry and Nawaz, 2008*) and Sheffield (Gaston et al., 2005a*; Smith et al., 2006*; Thompson et al., 2005*) in the United Kingdom and the Flanders region in Belgium (Antrop and Van Eetvelde, 2000*; Casaux et al., 2007*; Gulinck and Wagendorp, 2002*; Hermy and Cornelis, 2000*).
The Flanders study region, in particular, forms a specific type and good example study area in the sprawl/fringe/semi-urban area context. The northern part of Belgium is one of Europe’s most densely populated areas with 443.3 inhabitants per square kilometer in 2003 (FOD Economie, 2008), contains some important European cities like Ghent, Antwerp and Brussels1, and still has an important agricultural production. In contrast with other European countries like the Netherlands, the United Kingdom, or Germany, Belgium (and thus Flanders as well) had a very liberal spatial planning policy since the 1950s, which resulted in the fact that the majority of the countryside of Flanders has a somewhat urban character. Especially the central part of Flanders now can be described as a very hybrid matrix of countryside, containing a few big cities, some towns, many smaller settlements and this all connected with very dense infrastructure networks, not seldom accompanied by vast linear areas of ribbon-building. Hidden behind these built-up structures, a considerable acreage of agricultural and natural open space (as opposed to urban ‘closed’ space) is still present, be it highly fragmented. With these properties combined, the Flemish landscape outside the city limits in many cases resembles neither countryside, nor cityscape but in fact something ‘in-between’: large areas of Flanders can thus be seen as a good example of a ‘semi-urban area’. Flanders is not a unique case; there are other regions with apparently the same characteristics, for example the Lisbon province north of the city of Lisbon, Portugal, the Veneto province in Italy, the Département Nord in the North of France, or the Mediterranean coastal regions of France and Cataluña, Spain.