Abstract
Brucella infects marine mammals over broad regions as evidenced by positive serology and as confirmed by polymerase chain reaction (PCR) and culture methods. However, disparities in serological tests and the inability to culture has left many unanswered questions about its prevalence in Hawaiian monk seals (HMS; Monachus schauinslandi). Brucella isolates from marine mammals appear to be comprised of species distinct from Brucella spp. found in terrestrial mammals. While Brucella spp. found in terrestrial animals are pathogenic and often cause reproductive failures, the effects of Brucella spp. on marine species are only recently becoming known. Evidence of Brucella's role in the population dynamics of marine mammals includes reports of placentitis, abortions, and meningoencephalitis. None of these have been linked to positive Brucella serology in HMS.
Investigations of the seroprevalence of Brucella spp. in HMS began in 1997. Initial correlations of life history data to positive Brucella serology do not correlate with depressed survival or lower reproductive rates, but the limited investigation to date is based on a small sample size and limited data as well as many other mortality factors.
Serum was collected from 117 HMS at French Frigate Shoals (7), Laysan (49), Lisianski (17), Pearl and Hermes Reef (4), Midway Atoll (23), and Kure Atoll (17) between 1998 and 2002. One animal from Midway and one animal from Laysan were sampled on two different occasions. Samples were collected from apparently healthy animals. Serum samples were split and 119 subsamples were sent to both the National Veterinary Services Laboratories (NVSL) in Ames, Iowa and the Canadian Food Inspection Agency (CFIA) Animal Diseases Research Institute in Ottawa, Ontario, Canada. The NVSL tested for the presence of antibodies to B. abortus using the buffered acidified plate antigen presumptive test, particle concentration fluorescence immunoassay, rapid automated presumptive, tube agglutination, plate agglutination, and fluorescence polarization assay (FPA). The NVSL has no definitive interpretative guideline for classifying HMS serum using these serological tests developed for cattle. However, the National Marine Fisheries Service, Pacific Islands Fisheries Science Center, Marine Mammal Research Program (MMRP) has established the convention, based on laboratory and other expert advice, that serum tested positive to three or more of the tests is classified as a positive result. Less than three positive results is considered negative for the suite of assays used at this laboratory. This poster illustrates a comparison between the FPA performed at the NVSL and the CFIA, as well as results between the CFIA FPA and samples that were considered positive from NVSL.
Comparisons were made between the FPA assay at the NVSL and CFIA, where 6/119 (5.0%) samples were positive at the NVSL, while 13/119 (10.9%) were positive at the CFIA. The six positive NVSL FPA samples were also positive on the CFIA FPA assay but the CFIA FPA also identified an additional seven samples as positive that the NVSL FPA did not.
When using the MMRP convention, 12/119 (10.1%) samples were positive at the NVSL, while13/119 (10.9%) were positive for the CFIA FPA. Of the 12 samples that were positive at the NVSL, 9/12 (75%) were also positive for the CFIA FPA. Of the 107 samples that were negative at the NVSL, 4/107 (3.7%) were positive for the CFIA FPA.
The disparity between the FPA test performed at the NVSL and CFIA greatly limits confidence in the assay as being predictive for Brucella antibodies in HMS, and further work is necessary to resolve the differences. The 75% agreement between the positive samples and 96.3% among the negative samples when comparing the NVSL results to the CFIA FPA results is improved and may be predictive, but differences between the laboratories are apparent. Continued development of isolation techniques and PCR identification, as well as further serological studies, are needed to elevate confidence in these Brucella tests.
Acknowledgements
The authors would like to acknowledge the following for assistance with sample collection: A. Aguirre, G. Antonelis, J. Baker, S. Canja, M. Donohue, M. Craig, K. Cheves, D. Dick, B. Ryon, W. Sentman, M. Shaw, B. Stewart, C. Vanderlip, P. Yochem, the MMRP field camp personnel, and anyone else that we may have forgotten. A special thanks to H. Rhinehart and C. Maxwell.