Title: Pond water color and fish-eating birds: Improving the model? Authors
|Perschbaucher, Peter - UAPB|
Submitted to: Book of Abstracts Aquaculture America
Publication Type: Abstract Only
Publication Acceptance Date: September 3, 2007
Publication Date: February 9, 2008
Citation: Radomski, A.A., Ludwig, G.M., Perschbaucher, P., Thomson, S.J. 2008. Pond water color and fish-eating birds: Improving the model [abstract]? Book of Abstracts Aquaculture America. p. 19. Technical Abstract: Cormorant and other fish-eating bird numbers and activity were quantified from observation towers at each 3 spatially isolated catfish farms throughout the over-wintering months in Arkansas. During the first phase of the study, treatment ponds received AquashadeTM, a U.S. Environmental Protection Agency approved dye for use in aquaculture. The treatment was applied to 4 ponds at each of the 3 farms and replicated over 2 years. During the next phase of the study, we expand the predictive model of whether fish-eating bird use and avoid catfish production ponds by addressing the ultraviolet (UV) spectrum. During the initial phase of the study, we measured (1) physical (size, length, width, depth, surface condition, secchi depth), (2) biological (fish size and stocking density, algal species composition and density, and chlorophyll a), and (3) physiochemical (pH, temperature, salinity, dissolve oxygen, total phosphorus, TAN, nitrite, nitrate, and hardness) variables within catfish ponds every 10 days to determine pond dynamics. In conjunction, fish-eating birds (namely, double-crested cormorants, great blue herons, and great egrets) were quantified over the duration of the study. Preassessment of pond use by birds were classified as 'preferred' or 'avoided'. Assessment of the visible light spectrum of pond color was conducted using pond images obtained by low-altitude aerial digital video. The UV spectrum of randomly selected preferred and avoided ponds was measured using a fiber optic spectrometer. Computer software programs generated digital numbers to classify each pond based on the red, green, and blue wavelengths. Since this initial phase of the study indicated that no single variable was able to predict >60% of bird use of the pond and even though some pond attributes were slightly better at predicting those fish ponds with greater bird use, the best predictive model yielded could not be supported with biological reasoning. Results of the UV model will be added to the visible spectrum analysis and these findings will be reported.