APPENDIX A 



DATA PROBLEMS 



The even- and uneven-aged plots used in this study provide an unusual modeling opportunity 

 for western species. Unfortunately, their use also posed numerous modeling problems: 



1. Narrow geographic and site range . --All of the uneven-aged plots are situated within a 

 square mile tract, and the even-aged plots are about 1 mile away. The uneven-aged plots have 

 a narrow range in site and, because of this, site index as an independent variable may not be 

 significant for predictive purposes. In addition, the even-aged plots have an average site 

 index higher than uneven-aged plots. IVhile the differences in site index may be insignificant 

 for the uneven-aged plots, the difference between the uneven- and even-aged plots may well be 

 significant. All this could have contributed to the problems encountered while trying to 

 incorporate site index in the various component models of the simulator. 



Fortunately, one mitigating factor for these problems is that all the plots (both even- 

 and uneven-aged) occur on the same habitat type, Pinus ponderosa/Festuca arizonica (ponderosa 

 pine/Arizona fescue) , and it is believed that the plots and their site indices are representa- 

 tive of the type. Therefore, the simulator developed in this study is probably applicable to 

 that habitat type, at least in northern Arizona. 



2. Changes in diameter measurement limits . --Ideally, the smaller the lower diameter 

 limit, the better. Large, lower diameter limits ignore too much of the stand that is important 

 in understanding and modeling stand dynamics. The changing of lower limits on the Fort Valley 

 uneven-aged plots caused problems and raised several questions. A lower diameter limit of 7.6 

 inches would be necessary to utilize the full-time range of the data. This would provide six 

 to nine 5 -year growth periods of data and would incorporate cuttings at several times. It 

 would, however, also ignore about one important third of the stand--all trees under 7.6 

 inches . 



Conversely, a lower diameter limit of 3.6 inches would restrict the data to only the 

 early time periods. This would provide three to four 5-year growth periods, but it would also 

 eliminate many of the cuttings. I decided to use only the early data because: (a) a larger 

 portion of the stand would be modeled, and (b) by including more of the total stand, the 

 possibility of meeting objective number 2 would be increased. 



Because of the lower diameter limit of 3.6 inches for the uneven-aged stands, all data on 

 the even-aged plots below 3.6 inches were ignored. Several plots in the Taylor Woods study 

 have many small trees and, as a result, much of their structure was eliminated. A greater 

 portion of the number of trees in the total stand is usually located in the 0.0- to 3.5-inch 

 class of uneven-aged stands than in most even-aged stands, with the exception of very young 

 even-aged stands. While a lower diameter limit of 3.6 inches is an improvement over 7.6 

 inches, it is still far from the ideal solution of modeling the total stand. 



3. Intrusion upon uneven-aged plots . --Unfortunately, a major paved highway runs through 

 pieces of all the uneven-aged plots, and this could affect their stand dynamics. Not only 

 would these disturbed subplots have a serious "edge effect," but the two remaining subplot 

 pieces would not have the same competitive interaction as occurs on the undisturbed subplots. 

 The number of subplots seriously affected by the highway are given in table 21. 



46 



