40 BULLETIN U02, U. S. DEPARTMENT OP AGRICULTUEE 
experience that the larger the tree the more cones it produces. A 
30-inch tree on the average produces four times as many cones as 
a 20-inch tree. But this is not the only phase of the matter to be 
considered. Using the curved figures for cones per tree and average 
volumes in board feet for trees of different sizes, the average in- 
vestment in board feet per cone produced has-been computed for 
each diameter. This indicates, remembering that finality is not 
claimed for the conclusions, that beyond a diameter of 18 inches the 
number of board feet per cone increases rapidly. It takes 3.8 board 
feet to produce a cone on 18 and 20-inch trees, and nearly twice as 
much, or 6.5 board feet, on 30-inch trees; thus on the average an 
investment of 1,000 board feet in small trees should produce at 
least 50 per cent more cones than an equal investment in large trees 
of 30 inches in diameter. It seems clear that if trees are left solely 
for seed production those of a size having the largest cone production 
per unit of merchantable volume should be selected, which according 
to the data in hand are trees 18 to 20 inches in diameter. 
Similar detailed studies have been impossible for the other im- 
portant species, but it is reasonably certain that the relation of size 
to seed production is essentially the same for all species. 
But having determined the most desirable size for seed trees, 
the owner needs to know how many of these trees to leave in order 
to complete the reforesting of his land, and whether a diameter- 
limit cutting to the optimum cone-producing si^e will in fact secure 
for him the requisite number. 
NUMBER OF SEED TREES REQUIRED 
An extensive and typical national forest cutting in the yellow 
pine type, marked primarily with the idea of leaving only seed 
trees, rather than trees for growth, and now showing very good 
distribution of seed, will help in deciding the number of seed 
trees necessary* to give reasonable assurance of restocking. The data 
used are based on a careful 10 per cent cruise, or two strips per 40 
acres, made by A. E. Wieslander on over 2,000 acres of cut-over 
land. The average number of uncut trees 18 inches in diameter 
and over (that is, seed trees), using the ''forty" as a unit, is 3.9 
per acre, whereas the number most frequently found is 3 per acre, 
with 2, 4, and 5 trees also common. These seed trees range in 
diameter from 18 to 34 inches, with the great bulk between 18 and 
26 inches in size. Their distribution is excellent, for that is a 
specific requirement in a national forest cutting. The average 
volume of seed trees, 400 board feet, corresponds to an average 
diameter at breast height of slightly over 22 inches. 
These data are worthy of considerable emphasis because they 
indicate the method followed on national forest cuttings by experi- 
enced markers, and because the problem of interpreting the data 
is not complicated by trees left for growth rather than for seed, 
as is the case on cuttings in mixed t^^pes. Study of this area shows 
that the trees scattered seed everywhere, and it seems fair to con- 
clude that, given careful selection of seed trees for position and 
crown characteristics, three trees per acre are sufficient to insure 
restocking. 
