An experimental study of ontogenetic and seasonal changes in the temperature preferences of unfed and fed brown trout, Salmo trutta


Correspondence: J. Malcolm Elliott, Freshwater Biological Association, Far Sawrey, Ambleside, Cumbria LA22 0LP, U.K. E-mail:


  1. This study tests the hypothesis that the temperature preferences of brown trout, Salmo trutta, vary in relation to: (i) their mass and hence their age (0+, 1+, 2+, 3+); (ii) the season (spring, summer, autumn) and hence the acclimation temperature; and (iii) whether or not the fish are fed. The experiments were performed in still water in a 3-m-long artificial channel (water depth, 0.28 m). There was a temperature gradient from 25 to 0 °C along the channel.
  2. Experiments were for 0+ trout in May and for 0+, 1+, 2+ and 3+ trout in August and November. In each month, five trout of the same age were not fed for 2 days, and then one fish was placed in the section at noon and left for 24-h habituation. The next day, starting at noon, the position of the fish was noted every 2 h over 24 h, and the temperature at this position was measured (n = 13 readings). This procedure was repeated for each of the five fish (n = 65 readings for trout of the same age in each month) and then with five continuously fed trout of the same age as the unfed trout.
  3. For all trout, there was a diel variation in recorded temperatures at the observed positions with the lowest values in the day and the highest at night. This diel variation was slight for 0+ trout, but more marked for older trout, especially 2+ and 3+ fish. The overall mean preferred temperature decreased markedly with trout mass and age from 16.3 °C for 0+ unfed fish to 12.0 °C for 3+ unfed fish and from 14.5 °C for 0+ fed fish to 9.5 °C for 3+ fed fish. The mean preferred temperature was always lower for fed trout than for unfed trout.
  4. There was also an ontogenetic increase in the variation in individual temperature preferences around the mean value for both unfed and fed trout (CV% increased with fish mass and age). Neither the season (spring, summer, autumn) nor the acclimation temperature had any obvious effect on the mean preferred temperature.
  5. Therefore, the temperature preferences of brown trout were not constant but varied ontogenetically and whether or not the fish were fed. It is concluded from comparisons with earlier studies that the temperature preferences for fed juveniles (0+, 1+) were similar to their optimum temperature for growth rate, whereas the lower values for older fed trout were similar to their temperature for maximum conversion efficiency of energy intake to growth, thereby optimising growth efficiency rather than growth rate.