Spatial analysis of rural land ownership

Land ownership is an important determinant of landscape patterns. In rural southeastern Nigeria, private land ownership comprises the settlement, home -gardens and family-owned farmlands located in the vicinity of the settleme nt; the remainder of the land is communal. The focus of the study is on und erstanding rural land ownership in function of location variables. Particip atory rural appraisal, aerial photograph (1:6000) interpretation, and logis tic modelling were combined to spatially analyse private land ownership at Ikem, southeastern Nigeria. A binary logistic model for predicting private land ownership (PLO) correctly classified 76% of 1320 pixels and 72% of 330 validation pixels. The degree of coincidence between simulated and observe d PLO was fairly good (kappa = 0.559 +/- 0.002). The odds for PLO decreased with 90% per 100 m increase in distance from a track, with 50% per 500 m f rom a main road and with 34% per 500 m from the market. The driving forces for PLO were landform and distances to the market, roads and tracks. The mo del could be used to incorporate likely responses to changes in the determi nants into land use planning, impact assessment and scenario analysis. (C) 2001 Elsevier Science B.V. All rights reserved.