Campylobacter species carriage by Red-billed gulls fluctuated with seasonality. For instance, 250/0 of gulls sampled in May 1999 and 84.2% of gulls sampled in February 2000, were identified as carriers of Campylobacter species. Quantitative measurements identified bacterial loads of between 2.2 x 10² and 9.4 x 10³ CFU g⁻¹ of gull faeces. Genotypic techniques were used to determine relatedness between 76 clinical, 83 gull and three water Campylobacter isolates. 126 isolates were confirmed as C. jejuni or C. coli by PCR methods, using thermophilic Campylobacter-specific primers amplifying a 222 bp fragment of the 23S rRNA gene. Of these, 121 were shown to be C. jejuni and five C. coli based on cadF PCR. The remaining 36 isolates were presumptively identified as C. lari based on negative hippurate hydrolysis and nalidixic acid sensitivity. The amplification of the 23S rRNA gene, and amplification failure of the cadF and ceuE genes, supported identification. The pathogenic potential of each isolate was investigated through the detection of the virulence determinants flaA (flagellin structural gene), cadF (fibronectin adhesin gene) and ciaB (cell internalisation gene), at the transcriptional level. 96% of clinical isolates and 53.5% of environmental isolates generated PCR amplicons for all three loci. This suggested that some isolates of gull origin possessed at least some of the genes necessary for the initiation of campylobacteriosis. flaA-specific primers were applied to all samples, with 77.8% generating a 1.7 kb amplicon. RFLP analysis, using DdeI, distinguished 50 flaA profiles, with flaA type 2 being the most common type identified in this study. flaA RFLPs were diverse in isolates of environmental origin with 42% demonstrating unique profiles. In comparison, 19.7% of clinical isolates demonstrated unique profiles. SixflaA types were shared between both clinical and environmental isolates, these being types 2, 6, 7, 11, 16 and 35. This suggested either a common source of the pathogen or a bi-directional exchange of Campylobacter between human and gull populations. This technique was highly discriminatory with a D value of 0.96. PCR using gmhA-specific primers was applied to 35 flaA-grouped C. jejuni and C. coli isolates. 82.9% generated either a 0.9 kb or 1.6 kb band. Five gmhA types were generated using DdeI, the most common of these types being gmhA 1. gmhA 1 was shared between clinical and environmental isolates, supporting flaA clonality. PFGE, using SmaI, was applied to isolates belonging to the two largest flaA groups, 2 and 16, in an attempt to further assess relationships. This technique was discriminatory with a D value of 0.93. Additionally, C. jejuni survival in the soil environment was investigated. Persistence of culturability was largely a function of temperature. Cells were recovered by culture at 35 days in sand held at 4°C, at 7.5 days in sand held at 25°C, and at 54 hours in sand held at 37°C. This persistence, especially at decreased temperatures, suggested the potential for pathogen transmission from sand reservoirs into animal and human populations. The detection of C. jejuni DNA was also greatly affected by the microcosm temperature. At all three temperatures tested, DNA was detected after culturability was lost. By comparison, naked C. jejuni DNA demonstrated reduced temporal persistence. This suggested that C. jejuni may persist in a 'viable but non-culturable' state in sand that may contribute to the environmental cycling of the pathogen.