However, in a few cases the variation in the data is 

 so great that mean values may be visually different, 

 but not statistically different. For a more indepth 

 comparison of the two seeders and a discussion of 

 their performance, see Sloan (1990). 



Seed Mapping 



Overall, both machines sowed less seed than the 

 target rates (table 1). Mean seed densities were in- 

 consistent within species and for both seeders. In 

 fact, table 2 shows more consistency within nurser- 

 ies than within species. At each nursery different 

 persons calculated seed requirements and weighed 

 seed, and different tractor operators ran different 

 tractors with varying speed, precision, and consis- 

 tency. The hand-thinned plots were not mapped at 

 this stage because they were oversown. The seed 

 densities were not important so long as the seed 

 was distributed evenly over the entire plots. 



The average number of seeds can be misleading, 

 since spacing as well as density determines the seed- 

 lings' ability to compete for moisture, nutrients, and 

 light. Table 3 summarizes the seed spacing. Plots 

 sown with the Oyjord Seeder had more doubles 

 (seeds closer than half the average spacing) in seven 

 out of 13 tests. In two of the three tests at the J. W. 

 Tourney nursery, however, we found opposite re- 

 sults. Doubles did not differ between seeders at the 

 J. Herbert Stone Nursery and in one test each at 

 the J. W. Toumey and Coeur d'Alene Nurseries. 



Blank spaces were defined as any interseed space 

 greater than 1.5 times the average spacing. The 

 blank space results were similar to those for doubles 

 (table 3). 



The most important measurement is the mean 

 number of seeds that met the target spacing. A seed 

 with adequate spacing (a "single seed") is defined as 

 any seed that has a distance to seeds on either side 

 of 0.5 to 1.5 times the average spacing (Boyer and 

 others 1985). Neither the Summit Precision Seeder 

 nor the Oyjord Seeder was consistently more accu- 

 rate in placing seeds in these tests (table 3). 



Some of the seed placement performance by the 

 two seeders can be explained by seed characteristics. 

 The Summit Precision Seeder seemed to perform 

 better than the Oyjord when large, symmetrical, 

 and rounded seeds were sown. When small or 

 angled seed was used, however, the Summit sower 

 had problems placing one seed at a time. 



Of the seeds tested in this study, the spruces were 

 the smallest. White spruce was smaller than Engel- 

 mann spruce. Black spruce seed at the J. W. Toumey 

 Nursery was so small that it could not be success- 

 fully sown by the Summit Precision Seeder. The 

 western larch seed was also small. Ponderosa pine 



seed was the largest. For that reason it produced 

 the best spacing. Lodgepole pine seed was smaller 

 than the ponderosa, and red pine was smaller yet. 

 Jack pine had the smallest of the pine seeds. Douglas- 

 fir seed was large, but it was variable in shape. 

 The coastal variety planted at the J. Herbert Stone 

 Nursery was more triangular, while the inland vari- 

 ety seed at Coeur d'Alene and Lucky Peak was more 

 rounded. The Oyjord performed better than the 

 Summit with angularly shaped seed. Noble fir seed 

 was large, but it was also very resinous and sticky. 

 It often had pieces of wing still attached. This cre- 

 ated clumping problems in both seeders, especially 

 the Summit. 



Seedling Mapping 



Overall, the hand-thinned plots showed a lower 

 seedling density than both the Summit and the 

 Oyjord seeder plots (table 4) and, therefore, a 

 greater mean distance between seedlings. The ex- 

 ceptions came in the Engelmann spruce plots at 

 the Coeur d'Alene and Lucky Peak Nurseries, and 

 in the Douglas-fir plots at the J. Herbert Stone 

 Nursery where differences were not significant. In 

 the noble fir test at Wind River the Summit Seeder 

 produced a lower density and wider average spacing 

 than the hand-thinned treatment because of the 

 resin problem discussed earlier. 



Two tests were not included in the analysis of 

 seedling location, both at Lucky Peak: (1) the ponde- 

 rosa pine seedbed was established next to a shelter- 

 belt harboring seed-eating birds, which conducted 

 their own taste analysis during germination, and 

 (2) the Douglas-fir seedlot with an expected 95 per- 

 cent germination rate simply failed to germinate 

 at an adequate level. 



The hand-thinned treatment produced fewer blank 

 spaces and fewer double seedlings than the Summit 

 and Oyjord seeders. However, the hand-thinned 

 plots had a few blank spaces where seeds were not 

 sown or did not germinate. This shows that not 

 even the hand thinning was precise, putting preci- 

 sion seeding in perspective. 



The thinned plots had more well-spaced seedlings 

 (singles) in eight tests. In three tests at the J. W. 

 Toumey Nursery, however, the Summit Precision 

 Seeder matched or exceeded the number of singles 

 produced by hand thinning. With jack and red pine 

 the Summit Precision Seeder produced more singles. 

 With white spruce the difference was not significant. 

 The seedling maps showed even fewer differences 

 between the number of blanks, doubles, and single 

 seedlings than the seed maps. This demonstrates 

 that precision seeding can be totally effective only 

 when sowing seedlots with very high germination 



6 



