FISHERY BULLETIN: VOL. 70, NO. 3 



The present paper, therefore, owes a great 

 deal to Dr. Sette both in its conception and exe- 

 cution. He conceived that there was a relation- 

 ship between water temperature and fish dis- 

 tribution, and he stimulated the production of 

 charts to which we are able to refer in order to 

 establish the relationship of shad migrations to 

 water temperature. 



SHAD RUNS AND RIVER 

 TEMPERATURE 



As early as 1884, Marshall McDonald devel- 

 oped a theory that shad are restricted to a nar- 

 row range of temperatures and that the timing 

 of shad runs thus depends on when the water 

 warms. McDonald (1884, p. 599, 604, 605) 

 stated that shad 



. . . occupy an hydro-isothermal belt, or area, limited 

 by the temperature of 60° F. to 70° F. ; that they move 

 with this belt, i.e., as the season advances, into and 

 up the rivers. . . . 



... In the Savannah River they appear early in Jan- 

 uary, and in the Neuse River at a period not much 

 later than in the Savannah. In the Albemarle the im- 

 portant Shad seine-fisheries begin early in March, but 

 doubtless the fish are in the Sound some time before 

 that date; not, however, in numbers sufficient to justi- 

 fy the great expenses attendant upon the operation of 

 these large seines. In the Chesapeake Bay they make 

 their appearance in February, although the height of 

 the fishing season in its waters is during April and 

 May, and at a date somewhat later in the more north- 

 ern tributaries. In the Delaware, Connecticut, Mer- 

 rimac, and St. John (Nova Scotia) Rivers, Shad are 

 first seen at periods successively later as we proceed 

 farther north. The date of their first appearance in 

 any of these waters, however, varies from season to 

 season, the limit of such variation being from three 

 to four weeks. 



These irregularities in the time of the run into our 

 rivers, which cause so much perplexity and discourage- 

 ment to the fishermen, are, however, readily explained 

 when we keep in view what has been already said in 

 regard to the influences of temperature in determining 

 the movements of these fishes. 



McDonald pointed out that the shad run in a 

 particular river may be delayed if the water 

 cools. Although rising temperatures accelerate 

 the run, a sudden increase may also retard it. 

 The downstream migration of the juvenile shad 

 occurs when water temperatures fall to below 



60°F (15.5°C).' In support of his thesis, McDon- 

 ald provided a table of daily shad catches and 

 water temperatures in the Potomac River for 

 the year 1881. Shad were caught in water from 

 8.0° to 24.5°C. We performed a frequency anal- 

 ysis of these data which shows that the average 

 shad was associated with a temperature of 

 18.5°C. Ninety percent of the run took place 

 between 12.0° and 21.0°C. McDonald also sup- 

 l)lied a table showing monthly average water 

 temperatures for the St. Johns River, Fla., 

 in 1877 and 1878. He stated that shad first 

 appeared in November when the average tem- 

 perature was 63.5°F (17.5°C) and that spawn- 

 ing took place in February and March when 

 the average temperatures were 58.0° and 62.2°F 

 (14.5° and 17.0°C). McDonald provided no 

 quantitative data on shad in the St. Johns 

 River, however, and his thesis therefore appar- 

 ently rests on data from the one year and one 

 river in which both shad catches and water tem- 

 peratures were available. He gave no data to 

 support his statement on migrations of juveniles. 



Working in the field, one readily becomes con- 

 vinced as did McDonald (1884) that water tem- 

 perature affects the run of shad. This obser- 

 vation has also been made by Leim (1924), Tal- 

 bot (1954), and Massmann and Pacheco (1957). 

 The difl^culty has been to develop a quantitative 

 measure or prediction of the effect. Talbot 

 (1953), by means of linear regression, analyzed 

 the timing of the shad run in relation to river 

 flow and river temperature at Bonneville Dam 

 on the Columbia River. Temperatures and flows 

 are recorded there, and Talbot had available shad 

 counts made from 1938 to 1950 as the fish ascend- 

 ed the fish ladders at the dam. He concluded 

 that flow and temperature do affect the time of 

 entry of shad into the river, although the corre- 

 lation coeflicient of time of the run and temper- 

 ature was not quite significant at the 5''.^ level. 



Now 19 more years of data are available 

 (Table 1). The 32 years of data show a highly 

 significant i^artial correlation of temperature, as 

 well as a significant partial correlation of flow, 

 with the timing of the shad run. Of course, this 



' All temperatures used in developing this paper were 

 originally recorded in Fahrenheit. Our conversions to 

 Celsius have been rounded to the nearest 0.5°C. 



660 



