horizontal axis when 10 to 90 percent had hatched. The accumulative temperature 
curve is formed by adding the number of degrees the daily maximum temperature 
rises above 60° F., the theoretical point below which no hatching takes place. 
The legend explains the other curves. 
The earliest period when 10 to 90 percent hatched was April 26 to May 11, 
the latest July 5 to 23, a difference of 70 days or 10 weeks between the two 
periods. It should be noted that the early hatching occurred when exceptionally 
warm, dry weather occurred in March and April. The latest hatching took place 
when from March through June it was cool and wet. 
This difference of 10 weeks in the hatching period between years or regions 
means that small grain can be newly sprouted when the grasshopper infestations 
develop or it can be headed out and ready for harvest before the grasshoppers 
are big enough to do much damage. In eastern Montana in 1939, two areas 100 
miles apart had the same degree of infestation of sanguinipes. In one area 90 
percent of the sanguinipes had hatched by May 15 and all the newly sprouted 
grain was destroyed by this date. In the other area 9 percent had not hatched 
until June 8 or 24 days later. Here only slight damage occurred up to July and 
then migrating adults damaged 20 percent of the heads. The average daily maxi- 
mum air temperature for March, April, and May was 60.7° F. for the area of early 
hatching and 53.4° for the area of late hatching. 
What has been said of sanguinipes and shown in figure 1 is true of any 
economically important species of grasshopper. Late hatching of differentialis 
in south-central South Dakota permits the small grain to mature with little 
damage. When this is cut about August 1, differentialis, then in the last two 
nymphal instars and adult, moves from the grain into corn, flax, or other late 
crops, causing severe losses. 
Forewarming of the hatching period in any infestation enables early control 
of the infestation, which is more effective, efficient, and timely than waiting 
to protect a crop in answer to a farmer's plea for help. 
Method 
Because other anniyseel! demonstrated that the inclusion of rainfall gave 
no added support to the prediction, only daily maximum temperatures are used in 
this analysis. These temperatures measure warmth and depict the warming-up 
process in the spring, which in turn affects the hatching period. This pattern 
of weather is a continuous temperature variable in time marked by daily maxima 
and minima. 
By selecting equal periods of time such as 5 days, this continuous variable 
can be broken up into equal units of time. By taking the average daily maximum 
temperature for each 5-day interval, it is now possible to use these in a statis- 
tical analysis of the effect of temperature on hatching, wherein use is made of 
orthogonal polynomials. They become independent variables affecting a hatching 
date, the dependent variable when 70 percent or any percent of a given species 
has hatched. In this case it is the number of days after April 30 when 70 
percent of the end infestation had hatched. Thus the requirements for orthogo- 
nality are met and the computations of multipliers and regression coefficients 
can be carried out. 
3/ Shotwell, R. L. Unpublished data. 
=) 382 
