An investigation of the use of otolith microchemistry to discriminate reef fish populations and assess the movement of individuals.

摘要

The main objective of this thesis was to assess the degree of connectivity among populations of two highly abundant Caribbean reef fish (Stegasies partitus, Bicolor damselfish, and Haemulon flovolineatum , French grunt) at specific, but different portions of their life history; pelagic and demersal stages.; In Chapter 2 (Section A) I assessed the classification of individuals to the site and time they were collected. The classification of S. partitus to their collection sites, separated by as little as 5 km, was moderately successful, while investigations of the temporal variability revealed substantial variability at the scale of two weeks. This spatial and temporal variability in otolith microchemistry suggested that investigations of connectivity were possible, but would require frequent recalibration of chemical signatures.; Because the connectivity analysis of Chapter 2 relied upon the spatial variability in otolith chemistry, I investigated how the discrimination of populations could be improved in Chapter 3.; An assumption of the connectivity analysis used in Chapter 2 was that otolith elemental concentrations did not differ between fish of different life stages (i.e., larval/pelagic stage with its core chemistry versus juvenile/demersal stage with its edge chemistry). In Chapter 4, I assessed whether there was ontogenetic variability in otolith microchemistry by comparing the otolith chemistry of pre-hatch embryos to that of post-settlement juveniles collected at the same site and time. Results indicated that elemental concentrations of embryo otoliths were between 2 and 325 times greater than that of juvenile edge chemistry (and 2 to 94 times greater than water chemistry) for Mn, Zn, Ba, Ce, and Pb.; In Chapter 5 of Section B, I focused on the demersal stage of reef fish and whether otolith microchemistry could be used to discriminate H. flavolineatum caged in adjacent mangrove and coral reef sites in Belize and Bahamas. Significant variability in otolith trace elemental chemistry was detected among sites and habitats, which resulted in the classification of individuals separated by as little as 0.25 km (average correct classifications was between 68% and 85%).; In Chapter 6, I expanded the sampling of H. favolineatum (19 sites throughout Turneffe Atoll: 9 mangrove and 11 reef sites, separated by 0.8 to 20m kms) to assess the extent to which individuals could be correctly assigned to the sites from which they were collected when natural movements were permitted (i.e., in the absence of cages, see Chapter 5).; In Chapter 7, I discuss the findings of each of these chapters in the context of using otolith microchemistry in ecological investigations. (Abstract shortened by UMI.)