Sax and Gowen: Productive and Unproductive Apple Trees 297 



one trees. Of these fifty-one trees, 

 three or 6% are of type 1, while fifteen 

 or 29% are of type 3. A study of the 

 distribution of the two extreme types, 

 1 and 3, shows that they are in rather 

 definite areas of the orchard. Most 

 of the type 3 or unproductive trees 

 are located along the west side of the 

 orchard, especially towards the south 

 end of this area. The location of the un- 

 productive trees coincides very closely 

 with the high sandy portion of the 

 orchard. Clearly the grouping of the 

 unproductive trees is due to unfavor- 

 able soil conditions or exposure. The 

 type 1 or productive trees are grouped 

 largely in the west-central part of the 

 orchard, especially in the blocks in- 

 cluding rows 1-16, trees 9-16, and the 

 blocks including rows 25-32, trees 

 9-24. In these four blocks the per- 

 centage of type 1 trees varies from 

 25% to 44%. The grouping of pro- 

 ductive trees in definite areas can 

 only be attributed to unusually favor- 

 able soil conditions in these areas. ^ 



It is clear that the various types of 

 trees, productive and unproductive, 

 are due to a considerable extent to 

 soil heterogeneity. It is also evident 

 that an unproductive type of tree when 

 surrounded by a productive type, all 

 having a more or less common feeding 

 ground, cannot be attributed to unfa- 

 vorable soil. The behavior of such 

 trees indicates that the various types 

 of trees may be caused by one or both 

 of the other factors mentioned, i.e., 

 root stocks or bud variation. 



VARIABILITY IN PRODUCTIVITY DUE TO 

 ROOT STOCKS 



We shall consider next the influence 

 of the root stock on type and yield. 

 It is well known that seedlings grown 

 for root stocks are extremely variable 

 both in regard to morphological char- 

 acters and growth. In our seedling 

 orchard, the 586 trees which were 

 planted in 1911 vary in circumference 

 from two to eighteen cm. and the 



coefficient of variability was found 

 to be 32.07 + .69. Such great varia- 

 bility would be expected in the growth 

 of seedling apple trees since all apple 

 trees are more or less heterozygous 

 and are usually cross pollinated. 



So-called "standard stocks" may 

 vary greatly in growth; Hatton (c) 

 found about 14% of a block of "free 

 stock" seedlings to be distinctly 

 dwarfed and weak. About the same 

 range of variability in root stocks was 

 found in "free stocks," crab stocks, and 

 paradise stocks. The latter stocks are 

 generally considered dwarfing stocks, 

 but both the "free" and "Paradise 

 stocks were found to contain dwarfing 

 and free growing stocks. "Paradise" 

 stocks are apparently from different 

 sources and the variety is not a true 

 clonal variety. 



In our "stock and scion" orchard 

 we have over 400 trees of ten varieties. 

 These trees were worked on "French 

 Crab" and Tolman Sweet roots in 1913. 

 An analysis of the data on this orchard 

 shows that in circumference of trunk, 

 in 1921, the trees worked on " French - 

 Crab" stock are more variable than 

 trees on Tolman Sweet roots. This is to 

 be expected, since "French Crab" 

 seedlings are from numerous varieties 

 and types of trees. The Tolman stocks 

 are a clonal variety and are less 

 variable than French Crab stocks. 



When the trees for the "stock and 

 scion" orchard were taken from the 

 nursery they varied considerably in 

 size. If these differences in size were 

 due to difference in the soil of the 

 nursery we would not expect these 

 differences to be permanent when the 

 trees were transplanted in the orchard. 

 The random planting of large and 

 small trees should smooth out differ- 

 ences in growth if soil is the only 

 factor involved. If, however, the vari- 

 ation in size of nursery trees is due to 

 the effect of root stocks or scions 

 then differences in size will persist in 

 the trees when set in the orchard. 



^ The soil in our Ben Davis orchard was also found to be heterogeneous when measured by 

 the test for soil heterogeneity proposed by Harris (2). For the average soil heterogeneity based 

 on the yields of the 881 trees for 1914-18, r = .40. The individual tree was used as the ultimate 

 unit and trees were grouped in a four by four fold manner for the combination plots. 



