Date Awarded


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Virginia Institute of Marine Science


Joseph G. Loesch


This research investigated the relationship between juvenile abundance indexes (JAIs) for anadromous blueback herring (Alosa aestivalis) and abiotic factors (river flow and temperature) during the spawning and nursery period in the tidal freshwater areas of James, Pamunkey, Mattaponi, and Rappahannock Rivers in Virginia. Accomplishing this objective required the evaluation of the JAI methodology, specifically the effect of phototactic behavior on diel changes in the surface availability to the pushnet sampling gear, and obtaining population dynamics information (hatch dates, growth, natural mortality) during early life history from otolith microstructure of samples collected in 1991 and 1992. Mean catch-per-unit effort (CPUE) at night was one to two orders-of-magnitude higher than daytime mean CPUE. Availability dramatically increased approximately 30 minutes after sunset when light intensity was 10&\sp{lcub}-2{rcub}& to 10&\sp{lcub}-3{rcub}& &u&E/m&\sp2&/s. After the occurrence of this isolume, consecutive catches were order independent and varied without trend. A corresponding change in availability of prey did not occur, indicating that juveniles migrate to surface water in a specific isolume and not as a response to prey movement. Phototropic research verified the appropriateness of the JAI sampling methodology. Otolith microstructure analysis indicated that later larval emergence, reduced relative abundance, depressed growth, and increased mortality were associated with low water temperature and high river flow from a major episodic event. Results suggest that high river flows during the early larval period increase turbidity and reduce prey visibility, potentially causing depressed growth and starvation of newly hatched larvae. No linear or nonlinear relationship, however, was evident between any of the seasonal and monthly abiotic factors and the annual JAIs for the period 1979-95. All correlation and multiple regression analyses failed to reject (P &>& 0.05) the hypothesis of no effect. While the statistical power to detect a strong effect was high (&>&0.80), the power to detect weak effects was low, perhaps because of limited sample size and the lack of major runoff events during the study period. Power analyses indicated that the detection of a weak effect with the observed variance would require 25 or more years of data.



© The Author