Abstract
Combining multivariable statistics and geostatistics with landscape metrics, we attempted to quantify the spatial pattern of urbanization in the city of Niamey, Niger. Landscape metrics provided local quantification of both landscape composition and physiognomy while the Principal Component Analysis (PCA) yielded a multivariable summary of the main source of landscape metrics variation across the city. We used the variogram (geostatistics) to analyze the spatial pattern of the PCA outcomes and to characterize the associated spatial scales of variation. In Niamey, the main urban structure corresponded to a gradient ranging from highly diversified, fragmented, and both wooded and built-up areas in the city center and along the Niger River, to less green zones gathering steel-roofed houses whose density diminished towards the periphery. This concentric structure centered on the Niger River clearly reflected the history of Niamey. PCA and geostatistics provided appealing quantitative estimates of spatial patterns, scales, anisotropy and intensity of urban structures. Although these different tools are known in landscape ecology, they are rarely used together. The present paper illustrates how they allow characterizing the marked spatial variation of the urban landscape of the fast-growing African city of Niamey (Niger). Such a quantification of the urban landscapes may be extremely useful for future correlative investigations in various fields of research and planning.