SPECIAL TECHNIQUES 715 



To illustrate the method, a brief description of its apjjlicalion witli respect to estimating a 

 spring breeding population, is given. It is assumed that as a basis for the estimate, the com- 

 partments comprising the area were worked four times, at weekly intervals, during the month 

 of April. Data from the daily notes regarding flushes, drumming logs, droppings and kills 

 are then plotted on individual outline maps. Each working is indicated separately by record- 

 ing the information in different colored inks. As a further aid, flushes are plotted by sex as 

 judged in the field, using the conventional symbols S and ? for male and female respectively 

 while those of unknown sex are indicated by ^ . All three symbols are drawn so that their 

 points indicate the direction of flush as recorded on the data sheets. The difficulty of judging 

 sex under field conditions has been insofar as possible allowed for*. 



Having prepared the "flush map", it is then analyzed for possible reflushes that were not 

 recognized as such in the field. This has been done by utilizing location, sex data, distance 

 from supposed original flush, whether flushed from ground or tree, direction flushed, elapsed 

 time and color phase (if recorded). These reflushes must be discounted. Territories are then 

 plotted on the basis of the contact locations plus information from drumming logs and drop- 

 pings. Individuals that are judged to have been contacted on successive days usually may be 

 separated from individuals that were flushed but once during the period on the basis of loca- 

 tion due to the territorial habits during the spring period. Sex distribution has been estimated 

 on the basis of drumming logs, subsequent nest and brood information from late spring and 

 summer records, and from the estimation of sex at the time of flush. 



Familiarity with the area has been a distinct advantage in making population estimates. 

 Topography and cover often influence the destination of a bird once it is flushed, and a knowl- 

 edge of these items aids in the elimination of possible reflushes. In this connection, it has been 

 the practice for two persons to estimate the population individuaUy and then to arrive at a 

 final estimate by compromise. 



The application of any system for determining populations completely hides the inherent 

 variability of the original data and prevents checking the reliability of the estimate statis- 

 tically. Realizing the existence of a possible error, the Investigation attempted to check the 

 system used in the following manner. 



Based on the assumption that the number of grouse flushed per unit of effort would vary 

 directly with the population, if other conditions were equal, the daily flushes by sections, 

 recorded on the Connecticut Hill area during April over a ten-year period were statistically 

 analyzed by multiple covariance. Effort (man hours), wind (miles per hour), and rain 

 (inches) were used as the items affecting the number of flushes recorded. As might be ex- 

 pected, both rain and wind caused a field crew to record fewer flushes per unit of effort. 

 When the number of flushes was adjusted for the yearly variations that occurred, the resulting 

 trend in flushes was analyzed by regression against the estimated population. The relation 

 was found to be satisfactory and the conclusion was reached that the estimation system largely 

 eliminated the variations that might have been caused by adverse weather when the data 

 were recorded. Therefore, if it is valid to assume that the number of flushes recorded per 

 unit of effort, adjusted for wind and rain, is representative of the population, the estimates, 

 with only a small degree of error, may be considered to represent a true picture of the chang- 

 ing population of grouse. 



During the summer season the system has worked nearly as well, since adult territories in 

 general remained the same and means of distinguishing the various broods have usually been 



» See Chapter H. 



