This PD is a joint production of the members of the PICA Tool Task Force working at Cemagref in Clermont-Ferrand (Cemagref) and Humboldt-University of Berlin (UBER). A second empirical application of PICA was conducted, the perception of the tool by policy experts and their preferred modes of interaction were assessed and the tool was further integrated into SEAMLESS-IF, focussing on conceptual, technological, and empirical aspects. The main objective of this second testing was to analyse if a PICA application could be carried out by an external expert who was not involved in the tool development process. In order to achieve a full application of Seamless-IF, PICA was applied to the implementation of the policy option EU Nitrate Directive' in Midi-Pyrenees. An integrative modeller with a background in institutional economics was hired to carry out the testing. This was done in close cooperation with INRA in Toulouse, which provided expert supervision as well as the necessary infrastructure (working space, computer, data access etc.) to support the PICA run. The time frame for the testing was defined as two month, from July to August 2008. The IM has run PICA on her own, on the basis of the initial information provided by the SEAMLESS-PICA team and the information she collected. Therefore, as part of the testing, the IM has chosen a methodological design consisting of quantitative and qualitative approaches to collect available (regional) information to come up with qualitative statements about the Nitrate Directive's institutional compatibility with the context of Midi-Pyrenees. At all time of the PICA run, the possibility to contact the PICA team was guaranteed to clarify questions and to provide practical assistance. These requests have been documented in detail to use them for further improvements of the presentation of PICA in the GUI and preliminary training materials as well as for qualitative improvements of the tool. The application of PICA to the implementation of the Nitrate Directive in Midi-Pyrenees further validates the filter function of the policy matrix in PICA Step 1. With regard to the methodology used for the application of the subsequent steps, this testing highlights the interest of including several experts' opinions in order to reduce the subjectivity induced by the choices of the PICA Integrative Modeller alone or by the integration of stakeholders' point of view. The testing also shows the capability of the tool to be run by an expert who was not part of the SEAMLESS PICA development team. The conditions needed for future successful PICA applications were further defined. In terms of competencies, a background in institutional economics proved to allow the external expert for a quick adaptation to the tool. The existence of local contacts was of great help for an application in an unknown regional (institutional) context. As a timeframe for application, with the two previous conditions fulfilled, two months and a half seem to be sufficient to obtain relevant results. With regard to the improvements to bring to the presentation of PICA in the GUI and the training materials for future 'PICA-Integrative Modellers', precise methodological guidance at each step of the procedure are needed for an easier uptake of the tool. With these improvements, PICA should be ready for end-users applications. Finally, the combination of the PICA results with the results of the CropSyst-FSSIM model chain application (Louhichi et al., 2008) shows the added value of taking into account an institutional perspective in SEAMLESS-IF ex-ante policy assessment. The PICA results can contribute to the interpretation and validation of the model results, often dependent on strong assumptions with regard to institutional aspects which, in reality, influence policy implementation and outcomes. In addition to the testing of PICA by an Integrative Modeller in Midi-Pyrenees, policy experts as potential end users of the procedure have been included in testing activities. Therefore, four meetings between representatives of the Cemagref and policy experts from the regional services of French Ministries of Agriculture and Environment (potential secondary users) have been used as interacting platforms, where specific PICA workshops have been carried out to explore the perception of the tool by policy experts and their preferred interaction modes with future PICA Integrative Modellers. The interactions have shown the great interest of these potential secondary users towards the application of PICA. Improvements in the communication about the procedure for this specific audience were identified. Regarding the future interactions with 'PICA-Integrative Modellers', policy experts seem willing to be actively involved in the applications of the tool. Finally, two versions of the PICA GUI (not delivered to the European Commission) are presented: an August 2008 version integrated in the SEAMLESS-IF GUI and a March 2009 version, linked to the S-IF database These versions are based on a) suggestions made in appendix 3 in D2.4.2, and b) common work with WP5 during the project on PICA-elements, as documented in appendix 11 of the PD 6.5.5.2. Major improvements and adjustments have been made since then. Besides visual improvements, this also includes the creation of a PICA ontology and linking PICA with the database. Screenshots of each version as well as suggestions for further improvements are included in this PD.