Historic, Archive Document
Do not assume content reflects current
scientific knowledge, policies, or practices.
March1976 Division of«j2iZa2^esear^ s D A FOREST SERVICE
RESEARCH NOTE RM- 312
(EST SERVICE
DEPARTMENT OF AGRIC
PSW FCRCST AND RANG!
MAY 1 1 1976
LIBRARY COPY
Direct Seeding Ponderosa Pine
on Recent Burns in Arizona
W. J. Rietveld and L. J. Heidmann1
The significant factors determining success of direct seeding are those affect-
ing seedling survival, particularly moisture stress and frost heaving. Because there
is little or no control of these factors, the results of direct seeding are extremely
variable, even on the most favorable sites. Direct seeding recent burns, even by
spot seeding, is of limited usefulness in the Southwest because of high cost and
variability of results. The method should be reserved as a flexible tool to promptly
regenerate only the best sites when planting stock is unavailable.
Keywords: Forest regeneration, direct seeding, frost heaving, ponderosa pine.
Research has demonstrated that planting
forest trees is the most positive way to start new
stands when and where needed, but planting is
expensive. Direct seeding has the potential of
being more economical and flexible than planting,
but is less reliable.
The best time to plant or seed is immediately
after disturbance. Regeneration problems become
more difficult as reforestation is delayed. Recently
burned areas present the fewest obstacles to direct
seeding, and constitute an excellent test situation
to determine the success potential of direct seed-
ing in the Southwest.
Habitat conditions for seedeating rodents and
birds are least desirable immediately following a
fire. A hot burn also destroys most of the herba-
ceous seeds, so new pine seedlings compete with
fewer grass and weed seedlings of their own age.
This study examined the problems and prospects
of direct seeding on recent burns.
1Plant Physiologist and Silviculturist, respectively,
located at Station's Research Work Unit at Flagstaff, in
cooperation with Northern Arizona University; Station's
central headquarters maintained at Fort Collins, in coopera-
tion with Colorado State University.
Literature Review
Lavin (1955) found that fresh burns in northern
Arizona, on which the vegetation was largely
destroyed and the soil surface covered with a layer
of loose ash, were good sites for seeding grasses.
The practice of seeding grasses after fires has been
curtailed in recent years, however, because the
grasses compete strongly with tree seedlings in
later ponderosa pine regeneration programs. Lar-
son and Schubert (1969b) found that competition
for soil moisture was the main factor in competi-
tion between grasses and ponderosa pine seedlings.
Pine seedlings survived and grew best on plots
completely cleared of other vegetation. Grasses
start root growth earlier, grow faster, deplete soil
moisture rapidly to lower levels, and are capable
of enduring more drought than the pines. Although
partial shade from grass and weeds favors germina-
tion of pine seeds, the later effects of competition
are detrimental (Pearson 1950).
Water stress during germination and initial
development of ponderosa pine seedlings is one of
the main reasons for poor success with direct seed-
ing (Larson and Schubert 1969a). A slightly nega-
tive (-3 bars) water potential stimulated germina-
tion, but potentials below -7 bars greatly depressed
germination. Seeds alternately wetted and dried
several times germinated as well as unwetted
seeds, and significantly better than seeds soaked
1
for 24 hours without aeration. Root penetration,
root dry weight, and cotyledon length decreased
with decreased water potential of the soil solutions.
Another significant finding was that seedlings
that germinated under stress conditions grew
poorly even if they subsequently were well watered.
Low soil water potential is usually the result
of the precipitation pattern, competition of herba-
ceous vegetation, or a combination of the two
(Pearson 1942, 1950; Heidmann 1969; Larson and
Schubert 1969b; Schubert et al. 1970). Even in the
absence of competing vegetation, there is no con-
trol over the precipitation pattern, and direct-
seeded pine seedlings are more susceptible than
planted seedlings to temporary drought.
Study Areas and Methods
The study was established on two spring burns
that occurred in successive years on the Coconino
National Forest, Arizona. The Kelly Burn con-
sumed 3,550 acres in April 1971, and the Rattle
Burn consumed 714 acres in May 1972. Although
only two burns were available, contrasting soil
and site conditions are representative of those
found on the Coconino Plateau in northern Arizona.
On the Kelly Burn experimental plot, site index
was 65 (Minor 1964) at 7,470 feet elevation; soils
were Kelly sandy loam derived from basalt and
cinders. On the Rattle Burn plot, site index was
98, and elevation was 6,760 feet; soil was Soldier
sandy loam derived from limestone and sandstone
parent materials.
Kelly Burn Experiment
Direct seeding methods tested on the Kelly
Burn were broadcast, broadcast and rake, and
seed spot. Viable seeds of local origin were sown
at rates of 2, 4, and 8 pounds per acre by each
seeding method. There were four replications of
the nine treatments; one replication consisted of
four circular 1-milacre subplots treated alike. The
broadcast treatment involved spreading seeds
evenly over the soil surface with no attempt to
cover them. In the broadcast-and-rake treatment,
seeds were worked into the soil with a garden rake.
Seed spots were shallow depressions made with a
hoe in which the seeds were covered uniformly
with 0.5 inch of soil and compressed. Four seeds
were placed in each spot; the number of seed spots
per milacre increased with the seeding rate. The
Kelly Burn was seeded on July 14, 1971, 2 days
-before summer showers began. Schubert (1974)
recommends that direct seeding in the Southwest
be done in late June or early July, just before the
beginning of summer rains, to minimize the period
seeds are susceptible to rodent and bird depre-
dation.
Rattle Burn Experiment
Because of lessons learned from the Kelly
Burn experiment in 1971, only seed-spot and
broadcast-and-rake treatments at rates of 4 and
8 pounds of seed per acre were tested on the Rattle
Burn in 1972. In addition, three frost-heaving
"retardants"— gypsum (calcium sulfate), sodium
tetraphosphate, and coarse sand— were tested for
their capacity to reduce frost heaving of seedlings,
and for their interaction with seeding method.
Coarse sand was applied as a 0.5-inch layer over
the seed spots only; it was impractical to cover
an entire broadcast-and-rake subplot. There were
8 replications of the 14 treatments, each repre-
sented by 1 rectangular milacre subplot.
On designated subplots, either gypsum or
sodium tetraphosphate were added to the surface
1 inch of soil at rates of 1 percent and 0.5 percent,
respectively. Seeds of local origin, but from a dif-
ferent seed lot than that used in the Kelly Burn
experiment, were sown by the prescribed methods
and rates on July 6, 1972. Summer rainfall began
10 days later.
The measures used to retard frost heaving
and application rates employed were those found
to be most promising and nontoxic in concurrent
studies by Heidmann and Thorud (1976). Gypsum
lowers the freezing point of soil water, while sodium
tetraphosphate disperses the colloidal fraction of
the soil. Coarse sand is thought to disrupt the
migration of water to the soil surface to form the
ice lenses responsible for heaving.
General
All plots were examined for emerging and
dead seedlings at weekly intervals during the ger-
mination period, and for dead seedlings at intervals
of approximately 1 month during 2 growing
seasons. Causes of mortality were noted when
possible.
The rate and completeness of germination
were incorporated in "speed of emergence," a
parameter adapted from Maguire's (1962) "speed
of germination." Here speed of emergence (SE) is
defined as: SE = 1 number new emerging seed-
lings/number days from beginning of rainfall.
Speed of emergence was calculated for each inven-
tory and summed over the germination period. A
higher value reflects earlier emergence of a large
number of seedlings. Percent stocking was also
calculated; milacre subplots supporting at least
one live seedling were considered stocked. On the
Rattle Burn experiment, emerging seedlings on
three plots were marked with color-coded nails in
an attempt to link the date of emergence and fate
of individual seedlings.
2
Results and Discussion
Kelly Burn Experiment
Germination.— Spot seeding resulted in earlier
and more complete germination, expressed as
speed of emergence and number of seedlings per
acre, than either the broadcast-and-rake or broad-
cast treatments (fig. 1). Emerging seedlings were
first noted in seed spots on August 5, 20 days
after the beginning of the summer rains. By Au-
gust 8, germination in seed spots (3,800 seedlings/
acre) was considerably more advanced than on
broadcast-and-rake subplots (1,350), which in turn
was far better than the broadcast treatment (300).
Differences in speed of emergence among the
three seeding methods were highly significant.
Differences in number of seedlings per acre between
the seed-spot and broadcast-and-rake methods
were not significant beyond September of the first
growing season, however. The broadcast-and-rake
and seed-spot methods were similar in terms of
percent stocking (fig. 2), but the broadcast method
still showed a very low success level.
The differences in speed of emergence and
number of seedlings per acre among the three
seeding rates were highly significant and linear.
The abundance of seedlings germinating early
and number of seedlings per acre increased directly
with seeding rate. However, the 2-lb/acre rate
resulted in an unacceptably small seedling popu-
lation, even with spot seeding.
The initial success of the seed-spot method,
as evidenced by the more rapid and complete ger-
mination on the Kelly Burn, can be attributed to
the uniform covering and the compaction of soil
over the seeds to facilitate transfer of moisture.
The slight depression of the seed spot also served
to trap rain water. In contrast, seeds broadcast
and raked into the soil are covered to variable
depths, often too deep or too shallow, and are
surrounded by loose soil which limits the transfer
of water from soil to seed. Broadcast seeds, which
depend on soil washing to cover them sufficiently
for germination to occur, may be inadequately
covered or washed away. For example, heavy
showers in late July 1971, caused some sheet ero-
sion which probably carried away some seed from
the broadcast-and-rake and broadcast-seeded sub-
plots.
Seed spot
Figure 1. — Number of seedlings per acre (averaged over
the three seeding rates) resulting from the seeding
methods tested on the Kelly Burn during 1971 and 1972.
Drought in late summer and early fall, and severe frost
heaving in late fall and early spring, were responsible
for the excessive mortality.
J ' A ' S ' 0 ' N ' D '
I97I
■f'm'a'
M ' J ' J
I972
A ' S 1 0 1 N 1
3
Seedling Mortality.— While seed germination
was generally good in the seed-spot and broadcast-
and-rake treatments, seedling mortality began
immediately and involved a number of agents.
Drought and frost heaving were the most devastat-
ing. By May 1972 there were only a few surviving
seedlings on the entire experiment (fig. 1).
Some of the first seedlings to emerge were
clipped by rodents or birds, and only their tiny
stubs remained to record their emergence. Clipping
was not extensive, however, and occurred for only
a short period of time.
Erosion and burial caused by heavy showers
also accounted for small losses during the germina-
tion period when seedlings were small and shallow
rooted. Partial burial of seedlings by soil washing
did not seem harmful.
Competition from a heavy growth of weeds
and grass contributed to some seedling losses to
drought during the summer when rainfall was
sparse for 1- to 2-week periods, and for heavier
losses during September and early October when
rainfall declined (fig. 1).
The greatest seedling loss, however, was
caused by frost heaving during late fall and early
spring (fig. 1). By July 1972, after 2 periods of frost
heaving and a severe spring drought, only 10 seed-
lings were alive on the entire experiment. These
were concentrated on a part of the study area
where a few dead snags provided partial shade.
All seedlings, regardless of age or method of seed-
ing, appeared to be equally susceptible to frost
heaving. Pine seedlings whose deaths were ascribed
to frost heaving either lay on the ground partly
uprooted, remained upright but considerably
raised, or were broken off at the root collar. Some
seedlings survived, although they were raised a
half-inch or more. Some seedlings listed as dead
from drought may have died from injuries inflicted
by heaving stresses on the root collar and root
system.
80 r
Broadcast
and raked
o>60
c
\sc
o
o
"</>
c
0)
o
a>
CL
Figure 2.— Percent stocking, based on a minimum of one
live seedling per milacre subplot, resulting from the
three seeding methods (averaged over the three seed-
ing rates) tested on the Kelly Burn. Overall poor survival
is attributed to drought, intensified by competing vege-
tation and severe frost heaving.
40 -
20 -
J 'A'S'O'N'D'J'F'M'A'M'J'J'A'S'O'N'
1971
1972
4
Rattle Burn Experiment
Applying sodium tetraphosphate to the soil
prior to seeding significantly depressed speed of
emergence and number of seedlings (table 1). Gyp-
sum was also inhibitory in some instances, but
the effect was variable and may have been due to
chance. Germination and number of seedlings
were highest on subplots receiving no chemicals
and on seed spots covered with sand. The sand
acted as a mulch during the summer, but most of
it was washed away during heavy rains in October
1972.
Depressed germination caused by sodium
tetraphosphate and possibly gypsum was unex-
pected since these compounds were found to be
nontoxic at the same concentrations in the labora-
tory (Heidmann and Thorud 1976). The soils were
similar in the two cases: a fine sandy loam on the
Rattle Burn compared to a fine silica sand used
in the laboratory tests. A possible explanation for
the observed toxicity may involve fire-induced
soil water repellency and its effect on soil cation
exchange capacity.
20 r
Germination.— The seed-spot method was clear-
ly superior in speed of emergence and number of
seedlings (table 1, fig. 3). Following the onset of
summer showers on July 16, numerous seedlings
were apparent in seed spots by July 25 compared
Table 1. --Speed of emergence (SE) 1 and number of
seedlings per acre resulting from the two seed-
ing methods and two seeding rates, tested on
the battle Burn in 1972 and 1973
Seeding method
and frost-
heaving retardant
Speed of
emergence
Seed 1 i n
(Nov. 1
gs/acre
, 1973)
4 lb/
acre
8 lb/
ac re
4 lb/
acre
8 lb/
acre
- - No. - -
SEED SPOT:
None (control)
0.360
0.895
4,125
8,125
Gypsum
.343
■ 371
2,500
3,125
Sodium tetra-
phosphate
.296
.263
1 ,625
2,625
Sand
.471
.860
2,750
8,875
BROADCAST AND RAKE:
None (control)
.064
.108
750
1 ,500
Gypsum
.035
.158
750
2,625
Sodium tetra-
phosphate
.017
.036
250
625
16
CO
"D
c
o
CO
? 12
0)
0)
0)
in
1SE = E number new emerging seedlings/number
days from beginning of rainfall.
Figure 3.— Speed of emergence and number of seedlings
per acre were much greater by spot-seed than by broad-
cast-and-rake methods on the Rattle Burn experiment
in 1972 and 1973. The 8-lb/acre seeding rate yielded
approximately twice as many seedlings per acre as the
4-lb/acre rate. By November 1973, there were 5.4 times
as many seedlings on seed spots as on broadcast-and-
rake subplots.
Seed spot (8 lbs.)
Seed spot (4lbs.)
Broadcast and rake (8 lbs.)
Broadcast and rake (4lbs?)
J 'a's'o'n'd'j'f'm'a'm'j'j'a's'o'n'
I972 I973
5
to only a few seedlings on broadcast-and-rake sub-
plots by July 31. At the end of 2 growing seasons,
spot-seeded plots had 5.4 times as many seedlings
per acre (with no frost-heaving control) as broad-
cast-and-rake plots (table 1). The 8-lb/acre seeding
rate yielded approximately twice as many seed-
lings, by either seeding method, as the 4-lb/acre
rate. Percent stocking quickly rose to 100 percent
on seed spots and stayed at that level through
2 growing seasons for the 8-lb/acre rate; the 4-lb/
acre rate dropped to 75 percent stocking (fig. 4).
The broadcast-and-rake method yielded a more
scattered distribution of seedlings— 62.5 and 50
percent stocking, respectively— after 2 growing
seasons.
Percent stocking can be misleading when not
used in conjunction with number of seedlings per
acre. It is a measure of seedling distribution. Many
stocked broadcast-and-rake milacres had only a
single live seedling, while most spot-seeded mil-
acres supported several seedlings.
Differences in germination between the two
seeding methods were primarily due to the presence
of hydrophobic (water-repellent) soils resulting
from the fire. Most of the water repellency ap-
peared to be located in the surface inch of soil. Thus
depressions made for seed spots penetrated the
layer, while raking did not. Water penetrated in
seed spots, like small holes in a sheet of plastic,
whereas excessive runoff was observed from
broadcast-and-rake subplots. Immediately follow-
ing a rain, dry soil was found at a depth of 0.5 inch.
Part of the difference in success of the two methods
may also be attributed to seed losses from erosion.
Best results from broadcast-and-rake seeding
were observed in locations where soil was con-
siderably disturbed. Apparently, mechanical soil
disturbance is needed after eliminating competing
vegetation and exposing mineral soil by fire.
100 r
80 -
\ Seed spot (8 lbs.)
\
Seed spot (4 lbs.)
\
cn
c
O
o
Vi
C
<D
O
v.
0>
Ql
"\ Broadcast and rake
\ (8 lbs.)
60 -
40 -
20 -
\
\
Broadcast and rake
(4 lbs.)
Figure 4.— Percent stocking suggests smaller differences
between the two seeding methods and rates tested on
the Rattle Burn in 1972 and 1973, but the seed-spot
method and 8-lb/acre rate were generally better. The
increases in stocking in 1973 are due to delayed germi-
nation.
J ' A'S'O'N'D'J'F'M'A'M'J'J'A'S'O'N
B7Z
I973
6
Seedling Mortality.— Differences in number
of seedlings and percent stocking at the end of
2 growing seasons were primarily due to differential
germination, since mortality was similar in both
seed methods (fig. 3). Mortality, not excessive in
either method, was due to clipping and drought
shortly after germination. More losses due to water
stress were also incurred during the fall drought
in September and October. A heavy snowpack
which developed early and persisted late into the
spring of 1973 protected the seedlings from frost
heaving, normally prevalent during the first winter.
Dead trees on the experimental area, which pro-
vided overhead shade to preserve the snowpack,
also reduced the severity of day-night thawing
and freezing cycles. Losses of 2-year-old seedlings
to frost heaving during the winter of 1973-74 were
small despite saturated soil conditions conducive
to heaving.
Comparisons of seedling survival in relation
to germination date was not valid because too few
seedlings germinated late in the season, compared
to those germinating early. For germination dates
within the period July 25 to August 16, represented
by adequate numbers of seedlings, survival dif-
ferences were not significant on June 1, 1973. Lar-
son (1961) found that seedlings germinating early
(July) had better survival the following summer
than seedlings starting late (August, September).
General
These results demonstrated that the outcome
of direct seeding may be extremely variable, de-
pending on the net influence of several factors
affecting seed germination and seedling survival.
The initial number of seedlings depends mainly
upon seeding method and rate, soils, seed preda-
tion by rodents and birds, and moisture stress.
While it is not too difficult to obtain satis-
factory seed germination, the principal problems
limiting success are in keeping those seedlings
alive. The seeding method has little effect on such
seedling mortality factors as rodent and bird depre-
dation, moisture stress (intensified by competition),
browsing, and frost heaving— which can totally
eliminate a sizable seedling population. For these
reasons, it is extremely difficult to estimate the
size of an initial population of seedlings needed to
yield a minimum number of surviving seedlings
after a specified period of time.
The cost of direct seeding on seed spots, with
a minimum rate of 4 pounds of seed per acre, is
comparable to planting. Assuming the operational
costs of spot seeding at the 4-lb/acre rate and
planting 680 seedlings per acre are similar, the
current per-acre costs for seed and planting stock
are $79 and $33, respectively.2 Considering the
cost comparability, planting is favored because
it avoids or seriously reduces the problems with
frost heaving, rodent and bird depredation, and
variable stocking. Containerized tree seedlings
can be raised in as little time as 4 months.
This discussion discourages direct seeding on
recent burns, in comparison to planting, at least
until solutions are found for the problems of poor
seedling survival and variable stocking. Direct
seeding, as applied here, is an artificial regeneration
method and should be distinguished from natural
seeding, which is encouraged by regeneration
cuttings and supplemental site preparation. The
latter method has good potential for timely, inex-
pensive regeneration, and is receiving more atten-
tion.
Conclusions and Recommendations
The spot-seeding method proved to be con-
sistently better than the broadcast-and-rake
method with regard to speed and completeness
of germination, and number of seedlings per acre.
Distribution of seedlings was similar. The broad-
cast method was a failure due to slow, incomplete
germination and to low numbers of seedlings.
Because of high cost and variability of results,
direct seeding has limited potential as a regenera-
tion method on recent burns. Seed should be care-
fully applied in spots at a rate of at least 4 pounds
of seed per acre. Only the best sites should be
selected for seeding; clay soils should be avoided.
If planting stock from seed of local origin (Schubert
and Pitcher 1973) is available, it should be utilized
on burns in preference to seeding. When direct
seeding fails, time is lost, and it may be necessary
to invest more money in site preparation prior to
planting.
2Personal communication with Dr. John A. Pitcher,
U.S. Forest Service, Region 3, Albuquerque, New Mexico.
Seed costs are high because Working Capital Fund over-
head costs are included. Using commercially collected
seed, the cost of 4 pounds of seed (about $32) is still com-
parable with planting stock on a per-acre basis.
Literature Cited
Heidmann, L. J.
1969. Use of herbicides for planting site prep-
arations in the Southwest. J. For. 67:506-509.
Heidmann, L. J., and David B. Thorud.
1976. Controlling frost heaving of ponderosa
pine seedlings in Arizona. USDA For. Serv.
Res. Pap. RM- , p. Rocky Mt. For. and
Range Exp. Stn., Fort Collins, Colo. [In press.]
7
Larson, M. M.
1961. Seed size, germination dates, and survival
relationships of ponderosa pine in the South-
west. U.S. Dep. Agric, For. Serv., Rocky Mt.
For. and Range Exp. Stn. Res. Note 66, 4 p.
Fort Collins, Colo.
Larson, M. M., and Gilbert H. Schubert.
1969a. Effect of osmotic water stress on ger-
mination and initial development of ponderosa
pine seedlings. For. Sci. 15:30-36.
Larson, M. M., and Gilbert H. Schubert.
1969b. Root competition between ponderosa
pine seedlings and grass. USDA For. Serv.
Res. Pap. RM-54, 12 p. Rocky Mt. For. and
Range Exp. Stn., Fort Collins, Colo.
Lavin, Fred.
1955. Seeding in the Southwestern pine zone for
forage improvement and soil protection. U.S.
Dep. Agric, Agric. Handb. 89, 52 p.
Maguire, J. D.
1962. Speed of germination— aid in selection and
evaluation for seedling emergence and vigor.
Crop Sci. 2:176.
Minor, Charles O.
1964. Site-index curves for young-growth pon-
derosa pine in northern Arizona. U.S. For.
Serv. Res. Note RM-37, 8 p. Rocky Mt. For.
and Range Exp. Stn., Fort Collins, Colo.
Pearson, G. A.
1942. Herbaceous vegetation, a factor in natural
regeneration of ponderosa pine in the South-
west. Ecol. Monogr. 12:315-338.
Pearson, G. A.
1950. Management of ponderosa pine in the
Southwest. U.S. Dep. Agric, Agric. Monogr.
6, 218 p.
Schubert, Gilbert H.
1974. Silviculture of Southwestern ponderosa
pine: The status of our knowledge. USDA For.
Serv. Res. Pap. RM-123, 71 p. Rocky Mt. For.
and Range Exp. Stn., Fort Collins, Colo.
Schubert, Gilbert H., L. J. Heidmann, and M. M.
Larson.
1970. Artificial reforestation practices for the
Southwest. U.S. Dep. Agric, Agric. Handb.
370, 25 p.
Schubert, Gilbert H., and John A. Pitcher.
1973. A provisional tree seed-zone and cone-crop
rating system for Arizona and New Mexico.
USDA For. Serv. Res. Pap. RM-105, 8 p.
Rocky Mt. For. and Range Exp. Stn., Fort
Collins, Colo.
8
Agriculture— CSU, Fort Collins