708 METHODS Ai\D TECHNIQUES, 



formities noted hctweeii the two seem largely attributable to a difTereiice in altitude of 1.500 

 feet. 



Each time an iniuialure or adult grouse was contacted, the field leaders recorded the weather 

 information for a second series of problems. These observations were in general terms*, tem- 

 perature being noted, for exam|)lp. as warm, nornial or cold. In this case the basis of compari- 

 son was the average conditions for the ])articular time of year involved. 



Realizing that the influence of weather was the result of the inter-relation of the various 

 constituents, appropriate statistical methods were utilized in evaluating these interactions. The 

 relationship of grouse mortalitv rates to such items as temperature and rain, was determined 

 l)v means of multiple regression". The results indicated, for example, the effect of precipi- 

 tation on grouse mortality, with the effect of temperature also taken into consideration. 



The weather conditions, as recorded in general terms by the field leaders, were utilized in 

 determining the relation between this factor and the use of various cover types. Such 

 records, when summarized by seasons and anaKzed by chi-square' made it possible to judge 

 whether the use of various types of cover tended to vary according to changes in weather as 

 recognized by the field leader 



Determining Predator and Buffer Relationships 



Determination of the importance of predators and buffers as influences affecting grouse 

 abundance was one of the most difficult problems facing the Investigation. In attempting to 

 establish the true relationships, data were accunmlated regarding ]ircdalor abundance, buffer 

 abundance, food habits of predators and predators responsible for grouse found dead and nest 

 destruction. 



Predator Abundance and Activity 



Practical methods of estimating actual predator populations were not known and did not 

 become available during the course of the Investigation. Adequate methods of determining 

 trends in activity, however, were devised for the more important ground predators in 1933. 

 These methods centered around the idea of utilizing fluctuations in the number of jiredator 

 tracks as an index of activitv. The idea of activity is emphasized since it a])pears that the 

 fox. for exam])le. produces tracks in relation to its food supply. Similar populations of foxes 

 |)roduced man\ more tracks when food was scarce than when it was plentiful. Thus it is prob- 

 able that a general increase or decrease in tracks, throughout a period of years, may be an 

 indicator of fox numbers, while variations from year to \ear more iicarix represent activity. 



Predator activitv trends were obtained during tiic winter period when a tracking snow was 

 present. The various tracks were recorded on the daily notes in a prescribed maimer. Since 

 the amount of field work varied between \ears. it was necessary to reduce each year's data to 

 a com(>arable base through the use of a slatisti<al tool, nudtiple regression. I'tilizing only 

 those (la\s when a tracking sno« I urie inch nr more) was on the ground and taking into 

 account the numiier of man hours |>cr scrlion and tiie mimber of times tiie section was <i)vered. 

 the number of tracks recorded wa> adjusted to a base conmion for liolh ty])c and amount of 

 effort. Thus accurate \carlv fluctuations in the abundatice of predator tracks were estab- 

 lished. . -• 



• S«c fifiiircii T.t til 7(i. 



A See HiscUMi'iii iif Ki-Kn-iminn jiriil CiirrrUlinn, p. 720. 



t See dincussion n( Chi-aqtiiri*. p. 719, 



