METHODOLOGY 



DATA COLLECTION 



Reconnaissance Data 



A reconnaissance of the vegetation was the first step toward meeting 

 the objectives of the baseline study. Although a classification seemed to 

 be the most desirable way to handle the vegetation to meet the objectives, 

 the techniques to be used for classification had not been chosen. Franklin, 

 Dyrness, and Moir (1971) have outlined a reconnaissance technique suited to 

 community classification of which this study was roughly modelled. They note 

 the objective is to collect maximum data in a minimum amount of time. 

 Lambert and Dale (1964) define efficiency as the optimization of information 

 for a given quantity of work and note that it is often more efficient to work 

 with a large number of samples of low information content than with a smal- 

 ler number of much more elaborate and and time consuming records. These 

 objectives also guided this reconnaissance. 



The reconnaissance data were eventually used for community type class- 

 ification and definition, estimation of species diversity, and site des- 

 criptions. Tnese data also contributed to selecting productivity sampling 

 locations, making species lists, and served as ground truth for aerial 

 photography interpretation necessary for vegetation mapping. 



Figure 2 displays a field data form used in this reconnaissance. An 

 explanation of the collection of data follows: 



Sampled areas. Investigators sampled a total of 552 plots, more than 

 half of which were selected prior to field work from aerial photographs. 

 These aerial photographs were either 1:20,000 or 1:40,000 scale panchro- 

 matic or 1:12,000 scale true color. Sample areas were selected to cover a 

 range of photographic signatures, sites and geographic areas. 



Investigators chose other sample areas in the field to assure sampling 

 of communities not distinguishable from aerial photographs. Thus sample 

 area location was both objective (though not random) and subjective. 

 Mueller - Dombois and Ellenberg (1974) advocate "subjective sampling with- 

 out preconceived bias." Pfister and Arno (personal communication) also 

 advocate this technique, while rejecting random and systematic methods of 

 sampling because of inherent inefficiency. 



Upon reacliing a sample area, a team of two investigators observed the 

 general species composition of the stand and randomly picked a plot lo- 

 cation. If both team members agreed that tne resulting plot location did 

 not differ significantly from the stand in general, they sampled the plot. 

 Sample plots were set out along contours to minimize the chance of sampling 

 accross different soil strata, or microclimatic gradients. 



