Academic and other research is producing increasingly large quantities of digital output, much of it irreplaceable, and there is a pressing need to maintain long-term access to this data. Not only is the quantity of data available growing in size, it is also becoming much more diverse and complex, which significantly complicates the issues around its preservation. Data grid middleware has proved effective in managing large quantities of data, but until now has been restricted in the facilities it provides for implementing digital preservation functionality and for managing the associated complex metadata. In this paper we outline an approach to implementing digital curation strategies in data grids based on the iRODS middleware, in particular by exploiting iRODS' Rule Engine, which allows complex processing to be integrated within data grids, and we briefly describe the prototyping that we have undertaken.