Similar results have been reported related to viral infectivity and these same physico-chemical characteristics using field water samples. In addition, biotic components including WZ8040 distributor aquatic invertebrates and microorganisms have recently been proposed as potential factors affecting virus removal or accumulation in the environment. Finally, viral persistence in aquatic habitats has been demonstrated to be a determinant for IA virus transmission dynamics in wild duck populations. Several studies have attempted to isolate or detect IA viruses from surface water in habitats utilized by waterfowl. In these studies, virus subtypes detected in local aquatic habitats reflected the current subtype diversity circulating in waterfowl populations. Considering that water-borne transmission drives IA virus dynamics in wild birds, one could expect that a strong selective pressure may exist for IA virus maintenance in aquatic habitats. To date, environmental persistence and replication in the natural host have not been documented for water-isolated viruses, limiting our understanding of viral fitness in ducks and aquatic habitats. In this study, we performed the complete genome sequencing and estimated the persistence in water and replication in duck of two IA viruses isolated from surface lake water in Minnesota, USA. In particular, we investigated: the relative abundance of the two subtypes in viruses isolated in wild duck populations during three consecutive seasons, at the same sampling site; the genetic relatedness of these environmental isolates to IA recovered from wild ducks in Minnesota and North America; the persistence of the two viruses in water under different laboratory-conditions ; and the ability of these isolates to replicate in experimentally infected Mallards. We discuss the significance of the results regarding the role of water-borne transmission in IA virus ecology and epidemiology in wild duck populations. Both the H3N8 and H4N6 subtypes that were isolated from surface lake water have been commonly detected in North American waterfowl. At our study sites, the prevalence of these subtypes was relatively low in wild duck populations during the years of detection; however, the limited number of viruses recovered during 2006 provided limited information on virus subtypes circulating in local duck populations during that year. Interestingly, both the H3 and H4 virus subtypes were the most prevalent the year following their isolation in lake water. Although circumstantial, this pattern is consistent with a possible inter-annual environmental persistence. Phylogenetic analyses suggested that both isolates were reassortants related to viruses circulating in wild ducks in Minnesota at the time of sampling. In addition, we provide evidence that surface glycoproteins were highly similar to others identified in viruses circulating in North American ducks both before and after their detection in Minnesota. Based on phylogenetic analyses, no similar viral strain were however identified the years following their isolation in lake water, in Minnesota. This result suggests that, for the two water isolates, local persistence in aquatic habitats over winter was unlikely.