BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.

Road development is an important proximate cause of deforestation and consequent biodiversity loss. The interaction between road extension and land use calls for an integration of road simulation into land change models to improve model performance. To date there have been few attempts to simulate road development to support land change modeling. The objective of this research is to develop a road extension model to simulate roads network expansion. This road extension model forecasts the location of individual roads at local scale and produces the given spatial pattern of road network at regional scale. Locating individual roads is divided into two consecutive steps, identifying endpoints and locati * ng routes. Spatial pattern is controlled through altering the local road growth process, specifically limiting the number and locations of endpoints. The road extension model was released in 2006 as a component of the Land Change Modeler for Ecological Sustainability in IDRISI 15.0 (the Andes Edition). The model was applied to simulate scenarios of different road placement strategies in lowland Bolivia. The model results showed that the simulated road network exhibited varying spatial arrangements under different pattern parameter settings. The result provides insights into the ecological consequence of human land use policies. It also calls for an empirical based pattern parameter specification to predict road expansion that maintains existing network patterns.

Economic viability of farming operations determines long term success of US agriculture. This research examines the relationship between farming viability and its contributing factors at county level, using Ohio as an example. An ordinary least square regression (OLS) and a geographically weighted regression (GWR) model are developed to examine the effects of mechanical and biochemical technologies, government payments, product diversity and farm size. The OLS model shows that all factors are globally significant contributors to farming viability. The GWR model further reveals that the input on biochemical and fuel and the product diversity exhibit significant spatial variation in their effects on farming viability. The spatial variation of local effects is analysed in relation to the different geographic conditions and farming practices across eco-regions. The research improves the understanding of the mixed effect of explanatory factors on Ohio farming viability. The results also demonstrate that GWR model is a promising tool to acquire localized understanding of interaction between human and natural systems. JEL classification: Q10

No abstract