THERMAL BEHAVIORAL RESPONSES OF 

 SELECTED CALIFORNIA LITTORAL FISHES 



Karl F. Ehrlich,' J. Myron Hood,^ Gerald Muszynskl^ and Gerald E. McGowen^ 



ABSTRACT 



Two horizontal temperature gradients were used to measure behavioral responses to temperature of 

 various life stages of 16 species of temperate marine fishes from southern California, and we offer 

 guidelines for standardization of collection, acclimation, preexperi mental holding, and testing condi- 

 tions. We classified behavioral responses to thermal gradients using eight experimental parameters: 

 initial, mean, modal, and final selected temperatures; range of selected temperatures; skewness; 

 kurtosis; and the degree of dispersion between individuals. We found four behavioral responses to 

 changes in temperatures with time: 1) immediate — no general shift in selected temperature with time, 

 2) fast — a shift in selected temperature over the first 2 h of the experiment only, 3) slow — a shift in 

 selected temperature over more than 2 h in the experiments, and 4) positioned — a wide range of 

 selected temperatures and a tendency to remain in a given position in the gradient until conditions 

 become extreme. Effects of preexperimental acclimation temperatures on thermal selection did not last 

 longer than 4h. One day of food deprivation resulted in lower selected temperatures and changed the 

 precision of selection and aggregating tendencies, although the direction of the change varied between 

 species. 



Over the last several decades, the behavior of 

 fishes with respect to thermal gradients has been 

 investigated by many workers (see review by 

 Coutant 1977). Most studies in the literature, 

 however, have dealt with freshwater or estuarine 

 species that occur in large part in eastern North 

 America. Little work has been carried out on 

 temperate marine species, which, except for inter- 

 tidal inhabitants, generally experience smaller 

 natural diurnal and seasonal temperature fluctu- 

 ations. 



Our two horizontal gradients and experimental 

 procedures were built on many of what appear to 

 be the best attributes of previous studies. Varia- 

 tion in reported experimental techniques makes 

 comparisons between studies difficult. We suggest 

 guidelines for method standardization in this 

 paper. 



Horizontal gradients were chosen because they 

 allowed experimentation with groups, permitted 

 demersal fish to be in contact with a bottom and 

 pelagic species to swim freely. This tj^pe of gra- 



^Department of Biology, Occidental College, Los Angeles, 

 Calif.; present address: Lockheed Center for Marine Research, 

 6350 Yarrow Drive - Suite A, Carlsbad, CA 92008. 



^Department of Mathematics, Occidental College; present ad- 

 dress: Department of Mathematics, California Polytechnic Uni- 

 versity San Luis Obispo, San Luis Obispo, CA 93407. 



^Department of Biology, Occidental College; present address: 

 University of Southern California, Institute for Marine and 

 Coastal Studies, University Park, Los Angeles, CA 90007. 



Manuscript accepted May 1978. 



FISHERY BULLETIN: VOL. 76, NO. 4, 1979. 



dient also allowed us to combine some of the best 

 attributes of spatial gradients (e.g., Norris 1963) 

 and temporal ones (e.g., Beitinger 1976). Further, 

 our study, in conjunction with a field survey, was 

 designed to assess the effects of temperature on 

 the distribution, behavior, and physiology of fishes 

 found in King Harbor, Redondo Beach, Calif. This 

 harbor, which receives the thermal discharge from 

 an electricity generating station, as well as cold, 

 upwelled water from the adjacent Redondo Sub- 

 marine Canyon, contains a highly diversified 

 thermal environment (Stephens 1972), including 

 many horizontal gradients. 



Our intent is to introduce a comprehensive ap- 

 proach to thermal response testing, including 

 equipment design, preexperimental and experi- 

 mental methodology and protocol, and to show 

 representative examples of behavioral responses. 

 We have examined to date the behavior of various 

 life stages from larvae to adults of 16 fish species 

 from 10 families. Comparative studies of labora- 

 tory and field results are in preparation. 



EXPERIMENTAL CRITERIA 



A wide range of experimental methodology 

 exists in the literature (see review by McCauley 

 1977). McCauley and Tait (1970) stated, "Compar- 

 ison of preference temperatures in the literature is 



837 



