UC Santa Barbara

Decisions about which places to conserve are based upon the geographic heterogeneity of three types of information: public goods or benefits, their vulnerability to threats, and the costs to avert those threats. The choice of public goods depends on the mission of the conservation organization (e.g., biodiversity, open space, cultural values, or farmland). For spatial targeting of conservation at the regional scale, practitioners must estimate the values of these types of information. The quality of the estimations will vary by the primary data used, the assumptions made, and the practitioner's technical ability to analyze complex data. This paper contributes to the growing literature by presenting a systematic evaluation of effect of the quality of the estimation on the cost-effectiveness of the set of sites selected for conservation based upon those estimates. The specific case study targets farmland for preservation from urban development in California's Central Valley where a new land trust was recently established to purchase conservation easements. In one analysis, we compared the cost-effectiveness of farmland benefits using our most sophisticated estimation procedures to those that ignored costs and/or potential loss (i.e., assumed they were equal among sites). Excluding information about the potential loss of resources caused only a slight decrease in cost-effectiveness. On the other hand, ignoring cost information was extremely inefficient. The second analysis compared the performance of the sophisticated estimate to increasingly simpler estimates, such as those that are representative of the methods used by many American farmland preservation programs. The simplification of the estimates caused a 5- to 20-fold decline in the benefits that could be retained for a given budget. To make more cost-effective targeting strategies accessible to farmland preservation programs, we recommend that researchers develop new spatial targeting tools to overcome obstacles in data processing. © 2010 Elsevier Ltd.