A virus causing chlorosis and leaf distortion in the Western Australian endemic legume Hardenbergia comptoniana was detected by biological indexing to Chenopodium quinoa and Nicotiana benthamiana. Enzyme linked immuno-sorbent assay (ELISA) using general Potyvirus antiserum and amplification by reverse transcription polymerase chain reaction (RT-PCR) with degenerate primers indicated that it was a species of Potyvirus. It was confirmed as an unknown member of the genus Potyvirus by comparing its coat protein sequence with those of other potyviruses. The name Hardenbergia mosaic virus (HarMV) is proposed for this new virus species. Isolates of HarMV were collected from 13 sites, covering much of the natural range of its host. An experimental host range was determined using nine virus isolates tested against plants from 11 species in three families. Its infectivity on three leguminous species important in agriculture (Lupinus angustifolius, L. luteus and Trifolium subterraneum) was established. The nucleotide (nt) sequences of the coat proteins (CP) of 28 isolates determined there was 24.1- 27.6% diversity with the closest known relative, Passion fruit woodiness virus (PWV). Studies of the nucleotide sequences of the CP showed that there was considerable intra-species divergence (mean 13.5%, maximum 20.5%) despite its relatively small geographical distribution and single known natural host. The observed broad diversity strongly suggests long genetic isolation and that HarMV evolved in the region where it was collected. An examination of its phylogeny showed that 28 isolates clustered into eight clades with high bootstrap support (6.2-20.5% inter-clade diversity). Isolates collected at locations distant to the Perth metropolitan area (Margaret River and Seabird) diverged more from isolates collected in the metropolitan area (15.4-21.1% nucleotide sequence diversity). This virus represents the first endemic species to be characterised from Western Australia. Differences in pathogenicity and symptoms induced on key host species were seen between isolates belonging to different phylogenetic clades. Phylogenetic analysis confirmed the inclusion of HarMV within the Bean common mosaic virus group of the potyviruses and also defined a previously unreported subgroup of six previously described Potyvirus species (Clitoria virus Y, Hibbertia virus Y, PWV, Siratro 1 virus Y, and Siratro 2 virus Y), from Australia, which is further evidence for a prolonged period of genetic isolation. Both in relation to detection of strains of HarMV, and considering the broader issues of biosecurity and parallel detection of plant viruses, a microarray based detection system was established. To optimise conditions for the development of microarrays for virus detection poly-L-lysine (PLL) coated microscope slides produced in the laboratory were compared to commercially produced PowerMatrix slides (Full Moon BioSystems). Variables tested for PLL slide production were: choice of printing buffer, probe concentration, method of immobilisation and slide blocking; and in particular the print buffer and immobilisation method had the greatest effect on the quality of PLL microarray slides. Slides printed on PLL surfaces in a high salt buffer (3x Saline sodium citrate) supplemented with 1.5M betaine and immobilised at 42oC overnight retained the highest amounts of probe DNA of the methods tested. Qualitative comparisons of the two showed more probe was retained on PowerMatrix slides which were also more reliable and consistent than the PLL slides. Probes were designed for eight different virus species and six distinct strains of HarMV to test the potential to use microarrays to distinguish between them. Probes were designed to detect potyviruses at the genus, species and strain levels. Although there was evidence of non-specific hybridisation, the Potyvirus array was used to identify six strains of HarMV by hybridisation to species specific probes. Additionally the array was used to identify three other species of Potyvirus: Bean yellow mosaic virus, PWV and Passiflora foetida virus Y, following amplification with polyvalent PCR primers. In further microarray tests, using labelled first strand cDNA of Potato virus X (PVX) and Potato virus Y (PVY) on an array, PVX was strongly detected in leaves known to be infected, but PVY was only weakly detected in infected leaves. Three methods of pre-amplification of virus nucleic acid before hybridisation to the array were investigated to improve the sensitivity of the assay. Two of the methods, Klenow amplification and randomly primed PCR, amplified the target virus; as confirmed by real time PCR. Of the methods tested only randomly primed PCR improved the sensitivity of the microarray. The best amplification method used genus-specific primers with adaptor sequences. This method when tested by real time PCR showed a 3.7Ct reduction for PVX and 16.8Ct for PVY. The microarray correctly identified both viruses. In this work the first virus (HarMV) endemic to Western Australia was identified, and microarray methods were developed both to identify HarMV and other plant viruses of economic importance. The microarray approach, with further development, may be applicable as a means of identifying incursions of new viruses in a biosecurity situation.