Liver lesions in demersal fishes near a large ocean outfall on the San Pedro Shelf, California.

The prevalence of toxicopathic liver lesions in demersal fish on the San Pedro Shelf, California was determined for a 15-year period (1988-2003). Fish livers were sampled at fixed locations as part of the Orange County Sanitation Districts (OCSD) ocean monitoring program. Histopathological examination of selected fish liver tissues was studied to determine whether the wastewater discharge had affected fish health. The prevalence of toxicopathic lesion classes neoplasms (NEO), preneoplastic foci of cellular alteration (FCA), and hydropic vacuolation (HYDVAC) varied among species and locations. For all species sampled, severe lesions occurred in 6.2% of the fish examined (n=7,694). HYDVAC (4.1%) was the most common toxicopathic lesion type followed by FCA (1.4%) and NEO (0.7%). HYDVAC occurred only in white croaker (Genyonemus lineatus), accounting for 84.8% of the toxicopathic lesions for this species. Prevalence of HYDVAC, NEO, and FCA in white croaker was 15.2, 2.0, and 0.7%, respectively. The prevalence of HYDVAC and NEO in white croaker increased with age and size but there was no sexual difference. A linear regression model was used for hypothesis testing to account for significant differences in fish size (and age for croakers) at the different sampling locations. This analysis showed that for HYDVAC there was no spatial or location effect for lesion rate or size/age of onset. For NEO, the model predicted that white croaker near the wastewater outfall may acquire these lesions at a smaller size/younger age, and at a higher rate, than at other sites. However, this result may be biased due to the unequal size frequency distributions and the low prevalence of NEO in white croaker at the different sampling sites. Bigmouth sole (Hippoglossina stomata) had a prevalence of FCA and NEO of 1.3 and 0.35%, respectively, but the prevalence and distribution of lesions was too few for statistical testing. There was no sexual difference for lesion prevalence in hornyhead turbot (Pleuronichthys verticalis) and the prevalence of FCA and NEO was 3.4 and 0.37%, respectively. FCA prevalence increased with size in hornyhead turbot and there were no significant spatial differences for lesion rates and fish size at lesion onset. Overall, consistent spatial differences for lesion prevalence were not demonstrated and highlight the analytical difficulties of detecting a possible point source impact when the effect is rare, correlated with the size/age structure of the population, and may be caused by exposure to unknown multiple sources. Thus, the usefulness of liver histopathology as a point-source monitoring tool is best applied to where the spatial scale of impact generally exceeds the home range of the target species.