134 



cotton stubs on air temperature gave good estimates of temperatures and 

 heat flux in the places that insects overwinter. Multiple linear regression 

 models that included other variables such as moisture, trash depth, or boll 

 diameters, usually improved the estimates. 



250. , and Bonham, C. D. 1970. Analysis of populations of the boll weevil 

 in one acre of cotton at Florence, South Carolina, in 1957-1959. J. 

 Econ. Entomol. 63: 1505-1510. 



Input of climatological factors (cumulative number of degree days above 55"?, 

 mean and minimum relative humidities, cumulative number of rainy days, and 

 cumulative rainfall) in 1 acre of cotton at Florence from 1957 to 1959, were 

 linear in time during the 3 years. The deviations from the linear inputs pro- 

 vide an interpretive background for analyzing major changes in populations of 

 Anthonomus grandis Boheman. Varying numbers of overwintering weevils in the 

 climatic regimes of the 3 years resulted in population increases depend°nt 

 upon the climatic inputs. Damage was poorly correlated with the number of 

 adult weevils. Late-season migration was triggered by lack of oviposition 

 sites in cotton squares. 



251. , and Bonham, C. D. 1970. Summer temperatures of the soil surface 

 and their effect on survival of boll'tfeevils in fallen cotton squares. 



*»••• J. Econ. Encomoi. 63: 1599-1602. ^ 



Mortality in populations of immature boll weevils, Anthonomus grandis Boheman, 

 in fallen cotton squares commenced when the summation of the index of time 

 (1-hour period) x temperature >38°C reached 60. All the weevils died when 

 the summation reached 550. Soil surface in Arizona cottonfields attains 

 a mean of 38*'C when the mean in temperatures are 40**C. Soil surface tempera- 

 tures of SS'C are common when ait temperatures are lov;er than 30°C and may 

 60°C when air temperatures are at maximums. Frequent and prolonged periods 



