On-farm agricultural biodiversity conservation has long been recognized as a fundamental resource to the maintenance of ecologic and economic functions. In this light, planned on-farm biodiversity is represented as an economic asset providing a flow of ecological services to direct use of farmers. In particular, crop-biodiversity, measuring diversity within and among wild and domesticated species, has been found to significantly contribute to the productivity of agricultural production through its effects on agricultural yields and incomes. Moreover, agricultural risk-management literature recognizes the safety role of diversification in the presence of market, climatic and other hazards; given that crops react differently to external shocks, risk-averse farmers will choose to allocate their land to different crops in order to face environmental risks. Among the factors affecting biodiversity, institutional failures at different scales are reported as potential cause of biodiversity loss both in developed and developing areas. Within the European context, subsequent CAP changes toward deregulation and adoption of environmental protection measures have been investigated with the aim to assess their potential effect on farm economic and ecological strategies; the present paper seeks to contribute to this strand of literature by empirically assessing the relationship between farmers? diversification choice and CAP decoupled intervention measures. Focusing on the CAP policy changes occurred since 2003, the paper asks the following research question: has crop-biodiversity been positively affected by the implementation of a system of payments progressively delinked from quantity and quality of production? In order to answer this question, the paper empirical strategy relies on farm-level information drawn from the RICA database which collect annual data on the structure and economic performance of Italian agricultural firms. The analysis is carried out on firms specialized in the cereal sector, as this is a sector that experienced much of the changes coming from the reform of the CAP intervention system. To that purpose, a biodiversity function is estimated along the observed sample covering the 2004-2007 period; biodiversity, measured as spatial diversity, is expressed as a function of a set of variables capturing socio-economic farm characteristics (inputs use, degree of agricultural innovation, age of farm manager), land and agronomic conditions (soil fertility, irrigation, altitude), and the variable measuring CAP support. Finally, we include a dummy variable for regional location in order to account for the effect of specific cultural and physical elements on farmers? production choice.
