The scientific, political and societal debate on how to prevent the erosion of the world biodiversity has largely concentrated on the effects of two alternative types of agricultural production, land sparing (intensive farming leaving more land for natural spaces) and land sharing (extensive farming richer in biodiversity). For a given production target, Green et al. (2005) conclude in favor of land sparing if the relation between biodiversity and yield is decreasing and convex. Indeed, with a shift to extensive farming, biodiversity then increases little on already cultivated land, while it decreases strongly on newly cultivated land. According to these authors and to Phalan et al. (2011a), available empirical data are in favor of a land sparing strategy. We extend the model of Green et al. (2005) by adding the reaction of supply and demand to prices. We show that extensive farming is more interesting than intensive farming for biodiversity if the relation between biodiversity and yield does not have a very high degree of convexity, if demand reacts to prices and if the profitability of extensive farming is lower. Therefore, we do not obtain the result of Green et al. (2005) when production results from market equilibrium. Besides, extensive farming is detrimental to consumers when their surplus is evaluated in a restrictive way as increasing in quantities consumed and decreasing in prices. Its effect on agricultural producers is indeterminate. Distinguishing food demand in vegetal and animal products, we show that extensive farming reduces mostly the size of the animal product market, which demand is more price elastic. Integrating demand for animal products increases the advantage of extensive farming towards biodiversity by increasing the price elasticity of demand. The opposite result is obtained when we consider the introduction of policies for the mandatory incorporation of biofuels in fuel. Indeed, the demand for biofuels varies little with prices, to the de