3.3. Regeneration Methods: Seed-Tree

Category: Silviculture Textbook Created: Friday, 06 July 2012 Last Updated: Tuesday, 04 February 2020


An even aged regeneration method in which a new age class develops from seeds that germinate in fully exposed microenvironments after removal of all the previous stand except a small number of trees left to provide seed. Seed trees are removed after regeneration is established (Adams et al. 1994).

  • seed_tree1_moorhead
Figure 3.3.1. A slash pine stand, with half the original stand still intact (right side) and the other half following the seed-tree cut (left side). Once the new cohort is established, the seed-trees will be removed. Photo Credit: David Moorhead, UGA, Bugwood.org
  • seed_tree2_fountain
Figure 3.3.2. A loblolly pine stand showing both the seed trees and the newly established cohort. The seed-trees will be removed after the new cohort is allowed to grow for one or two more years. Photo Credit: Michael Fountain


The seed-tree regen method is very common where natural regeneration is desired and the species to be regenerated are shade-intolerant and light-seeded.

Cuts Involved

The most common application of the seed-tree regen method involves two cuts.

  • The first is often referred to as the seed-tree cut. Its intent is to remove the majority of the trees in the stand to create a fully exposed microclimate for the establishment of seeds from a small number of high-quality mature trees left on site for this purpose. With the exception of leaving the seed-trees, this operation is very similar to a clearcut.
  • The second cut is often referred to as the final or removal cut, and occurs between 3 and 10 years following the seed-tree cut, depending on the species being regenerated. Its intent is to remove the seed-trees to release the newly established cohort. There may be significant value left in the seed-trees that can be captured with this harvest. This cut is necessary even if the seed-trees are not merchantable because they are too large for the mills, have been damaged by wind or lightning, or are simply not valuable enough to be worth paying a logger to set up on the site. If the seed-trees are not removed, they will suppress the growth of the new cohort.

A diagram depicting the two cuts in the seed-tree regeneration method described above.

Figure 3.3.3. The seed-tree regen method results in an even aged, single cohort stand upon completion of the removal cut. For a brief time between the seed-tree and removal cuts, the stand is two aged, with both the seed-tree cohort and the newly established cohort coexisting. However, because this is only a temporary condition, and the intent is the regeneration of an even aged stand, the seed-tree system is classified as even aged.
  • seed_tree3_fountain
Figure 3.3.4. A loblolly pine stand immediately prior to the removal cut. Photo Credit: Michael Fountain

Common Variations

The two most common even aged variants of the seed-tree regeneration method are the uniform seed-tree and the aggregated seed-tree. Both include the same cuts, but differ in the spatial arrangement of the seed trees.

A diagram depicting the uniform versus aggregated spacing of seed-trees.

Figure 3.3.5. Overhead view of uniform versus aggregated seed-tree distributions.

Uniform spacing is more common as it more effectively distributes seed across the stand with a minimum number of seed trees.

However, leaving aggregated patches of seed trees does offer several advantages.

  • Aggregated spacing may limit damage to the seed trees during the seed-tree cut.
  • If seed trees are susceptible to windthrow or lightning strikes, leaving multiple trees at each location will lessen the probability that a disturbance removes all seed trees from that area. Even if one or more seed trees are lost, there are several at each location, reducing the risk of a regeneration failure.
  • These small patches of large trees may also offer improved habitat for some wildlife as compared with uniform spacing.
  • Aggregated spacing can also improve operability during the removal cut and may minimize damage to the new cohort by lessening the proportion of the stand affected by skidding. If it is necessary to leave more seed trees than required to regenerate the stand primarily to ensure that the removal cut is a merchantable harvest, then leaving them in groups may be advised.
  • While not shown in the figure above, seed trees may also be retained in strips. This is more efficient for harvest operations. The strips can be placed so that they run perpendicular to the prevailing wind direction to maximize seed dispersal distances over cut areas, although the probability of windthrow should be considered before prescribing this. Strips can also be located along contours or up and down steep slopes, depending on the harvesting system being utilized.

For the two-aged modifications to the seed-tree regeneration method, see the sections on deferment and reserves.

Ecological Considerations (Silvics)

The seed-tree regen method creates a fully exposed microclimate for the establishment of the new cohort. Thus, it is only suitable for species that are intermediate, intolerant, or very intolerant of shade.

Seed must naturally disperse from the retained seed-trees. Therefore this method is best suited to species that are light-seeded, particularly those with wind-dispersed seeds. Heavy seeded species or species with seeds that are subject to high predation such as oaks or hickories are not well suited to the seed-tree regen method.

Retained seed-trees must remain on site for 3 to 10 growing seasons for most species to ensure adequate regeneration. It is important that seed-trees remain healthy and vigorous during this period to continue producing seed, and so they are merchantable at the removal harvest. Due to these constraints, species that are deep-rooted, and as a result relatively wind-resistant, are best suited to this method of regeneration.

Suitable Species

  • US South
    • Loblolly pine (Pinus taeda)
    • Slash pine (Pinus elliottii)
    • Longleaf pine (Pinus palustris)
    • Shortleaf pine (Pinus echinata)
    • Eastern cottonwood (Populus deltoides)
  • US North
    • Eastern white pine (Pinus strobus)
    • Red pine (Pinus resinosa)
    • Black birch (Betula lenta)
  • US West
    • Ponderosa pine (Pinus ponderosa)
    • Western larch (Larix occidentalis)
    • Douglas-fir (Pseudotsuga menziesii)
  • Interregional
    • Ash (Fraxinus spp.) (S,N,W)
    • Willows (Salix spp.) (S,N,W)

Economic Considerations

Compared to the clearcut method of regeneration, the seed-tree system requires that the removal of some volume (the seed trees) be deferred for a period of years. This entails some risk of loss of the seed trees due to windthrow, lightning, fire, ice, insects, disease, or changing market conditions. This also requires a second entry into the stand to remove the seed trees, which is less economically efficient compared to a single entry as the per-unit cost of production will be higher.

It may be necessary in some instances to leave more seed trees in the stand than are necessary for regeneration so that the final harvest will be merchantable. Familiarity with local market conditions is essential when determining the traits and number of seed trees to be left.

Retaining seed trees in the open for several years will increase their growth rate. However, this effect will be minimal if the seed-tree and final cuts are only a few years apart. While leaving large vigorous seed trees until the final harvest is essential to secure regeneration, it is important to know the maximum size logs that local mills can handle. Many Eastern US sawmills can no longer process logs larger than 28 inches in diameter, for example. Trees nearing the limits of the mills should not be retained as seed trees if it is unlikely that they will be merchantable at the final harvest.

If seed trees are desired to be sold as grade or veneer logs, first consider whether the species being managed is prone to epicormic branching. Epicormic branches are small branches formed from dormant buds along the stem in response to exposure to light. If seed trees develop too many epicormic branches, it can degrade the product class and dramatically lessen their value. The aggregated seed-tree method may reduce epicormic branching on high-value seed trees by leaving other, lower value trees around them to shade the bole. The disadvantage of this approach is that those lower quality trees may contribute their genetics to the next stand.

In that the seed-tree regeneration method relies on natural regeneration rather than planting, it may cost less for the landowner at the time of establishment. However, for species where improved seedlings are available, the new naturally regenerated cohort may not have the best form, growth rates, or disease resistance available. In the long-term, artificially regenerated stands can outperform naturally seeded stands even when the initial costs of seedlings and tree planting are considered. In regions such as the US South, where tree improvement programs have been operational for more than 50 years, the landowner may still reap some benefits of tree improvement even with the seed-tree system. If the previous rotation was planted in improved genetic stock, then the seed trees and the new cohort will still show genetic gain relative to unimproved phenotypes. However, they will likely be one or more cycles of tree breeding behind the currently available selections, and thus will likely grow more slowly.

Societal Considerations

The seed-tree system has several advantages over clearcutting from a societal perspective.

Aesthetically, seed-tree stands between the time of the seed-tree cut and removal cut appear to be open and park-like. They are very aesthetically pleasing to most people regardless of their knowledge of forestry. Following the removal cut the new cohort will usually be at least several feet tall, thus avoiding the appearance of a barren field of stumps.

Forest certification systems often limit the average size of clearcuts a landowner may implement in a given year. Certification also requires a minimum length of time between clearcutting adjacent stands, called the green-up requirement. These provisions usually do not apply to the seed-tree regen method, thus simplifying forest planning.

Silvicultural System Considerations

Selecting Individual Seed Trees

Seed trees should be selected to both provide sufficient seed to achieve full stocking in the new cohort and ensure the seed trees continued survival and viability at the removal cut. There are several key traits that should be used in selecting seed trees.

  • Trees with a large crown will maximize seed production. This is due to their greater vigor and large leaf area. The definition of a large crown varies by species. For example, in southern yellow pines a live crown ratio greater than 33% is required.
  • Trees should be at reproductive maturity with past evidence of good seed production. Past evidence of good seed production is the best way to assure future seed production.
  • Trees need to be windfirm. Increased taper is correlated to windfirmness, but may be a less desirable phenotype when managing for poles.
  • Trees should have desirable phenotypes, since they are providing seed for the next rotation.
    • Their form should be free of defects such as forking, sinuosity, or ramicorns (an excessively large and steeply angled branch that forms a large knot).
    • No evidence of serious insect or disease susceptibility should be present.
    • Good self-pruning is desirable for sawtimber, poles, or veneer.
    • Rapid growth rates increase volume production and shorten the rotation.
  • seed_tree4_moorhead
Figure 3.3.6. A slash pine stand with poorly selected seed-trees. The crowns are too small (live crown ratio < 33%), indicating that they will likely not have the photosynthate required to produce significant seed crops. Photo Credit: David Moorhead, UGA, Bugwood.org

Determining the Density and Spacing of Seed Trees

There are several scientifically sound methods for determining the density and spacing of seed trees, primarily for coniferous species. They vary in their data requirements and complexity, but all require advanced knowledge of the silvics of the species to be regenerated.

Method 1: Seed Dispersal Distance

This is perhaps the least accurate method available for determining seed tree density, but will do in situations where little more is known about the species other than its effective seed-dispersal distance. Seed dispersal can be measured by placing traps at varying distances from a clearcut edge if the adjacent stand is relatively pure and at reproductive maturity. Let's look at an example using shortleaf pine.

What you need to know

  1. Typical seed dispersal distance (Yocom 1968): 66-132 feet for shortleaf pine.
  2. Seedbed conditions.
  3. Windthrow and lightning risks.


    1. Calculate the area that will be seeded by each tree.
    2. Add appropriate safety factors based on seedbed condition and risks to seed tree loss.
    3. Convert this to the density of seed trees to be retained per acre.
      • A = π r2
      • A = π (66 ft)2 (use the minimum radius to get a conservative estimate)
      • A = 13,685 ft2 seeded per tree
      • 43,560 ft2/acre / 13,685 ft2/tree = 3.2 seed trees per acre
      • 3.2 x 3 = 9.5 seed trees per acre (conservatively assume the loss of up to half the seed trees to windthrow and lightning, and an unfavorable seedbed)
      • round 9.5 to 10 seed trees per acre

Selection of the safety factors will depend on your knowledge of these risks in your area, your understanding of how seedbed conditions and competition will affect regeneration success, and your experience with seed-tree regeneration for the species in question.

Once you know you want 10 seed trees per acre, imagine leaving them on a square grid. Thus, each seed tree will be left to cover 43,560 ft2/acre / 10 trees/acre = 4,356 ft2/tree. Assuming this area represents a square around the tree, take the square root of 4,356 ft2 to calculate the length of each side of the square. Thus, you arrive at a square spacing of 66 feet between each seed tree. This is how you would create guidelines to mark your prescription and implement it in the field.

Method 2: Relying on the Literature

Fortunately the seed-tree regeneration method has been applied for a century or more for many of the species it is suitable to regenerate. There is information readily available in the literature to estimate the number of seed trees required based on their size, and the spacing at which to leave them. These figures are typically developed using procedures similar to above, but also modified to ensure successful regeneration based on the experience of experts familiar with these silvicultural systems.

Table 3.3.1. Recommended retained seed-tree per acre by diameter for loblolly, slash, and shortleaf pines (Beaufait et al. 1984).  Following each density, the spacing between trees that will achieve that density is given.
dbh Loblolly or Slash Shortleaf
10 in 12 TPA, 60 ft 20 TPA, 47 ft
12 in 9 TPA, 69 ft 14 TPA, 56 ft
14 in 6 TPA, 85 ft 12 TPA, 60 ft
16+ in 4 TPA, 104 ft 12 TPA, 60 ft

Method 3: Cone Counting

This is the most reliable method of determining seed tree density based on more data than the previous two methods. Cones of southern yellow pines take approximately 2 years (24 – 26 months) to develop and release their seed. To accurately estimate the potential of a stand or a number of retained seed-trees to establish a new cohort in the next year or two, it is possible to count cones as they form (Croker and Boyer 1975). Cone counting is of particular importance in longleaf pine due to its infrequent and unreliable seed crops (Beaufait et al. 1984), but may be done in any of the other southern yellow pines. There are two key periods during a rotation of a naturally regenerated even-aged stand when it may be desirable to count cones:

      • Prior to implementing the establishment cut in a seed-tree or shelterwood regeneration method.
      • Following the establishment cut and prior to final cut in either regeneration method.

The best time of year to count 6-month-old cones at end of branches is when they are exposed in April – May (Croker and Boyer 1975). These cones will not release seed to regenerate the stand until after the second growing season (approximately 18 months later). Thus, these counts tend to be highly unreliable, but may be used to indicate a future crop failure. Counting 18-month-old cones among the foliage of previous year’s growth yields reliable estimates of seeding more than 90% of the time (Croker and Boyer 1975). These are cones that will release seed over the coming winter to regenerate the stand the following growing season.

Table 3.3.2. A timeline of cone development for loblolly pine as modified from Trousdell (1950).

A table showing the two-year development period of seed cones in southern yellow pines.

The general procedure for cone counting is:

      1. Select a sample tree (seed-tree that will be or has been retained).
      2. Standing approximately one tree height away from the tree, use 8-power binoculars to systematically count all 18-month-old cones on the tree. Keep your feet in the same place while doing this.
      3. Record the data and repeat this procedure on approximately 50 trees to achieve adequate accuracy.
      4. Double the cone count on each tree to account for underestimation of cones obscured by branches, foliage, and the bole (Yocom 1968).
      5. The data collected above will be in the units of cones/tree. The desired data that needs to be calculated is either 1) trees / acre to be retained, or 2) seeds / acre to determine if an establishment cut should be made in the next year. Calculate each of these as follows:

Equations that convert cones per tree to either trees per acre or seeds per acre using ratios equal to one.

To adequately regenerate a fully-stocked stand, 50,000 seeds per acre are required for longleaf pine (Croker and Boyer 1975). If it is impractical to count the number of sound seeds in sample cones, then the following guidelines from Croker and Boyer can be used to estimate seeds / cone (1975).

      • 50 seeds / cone in good year (risky estimate, can result in failed regeneration)
      • 35 seeds / cone in average year (typical estimate)
      • 15 seeds / cone in bad year (conservative estimate)


  • seed_tree5_fountain
Figure 3.3.7. Once the characteristics and spacing of seed trees is known, adequate marking assures that the prescription is correctly implemented. Photo Credit: Michael Fountain

Marking of seed trees should be sufficient to ensure that they will neither be unintentionally harvested nor damaged during the seed-tree cut. It may be necessary to mark a band around each seed tree, although if the marker knows the landing location and direction the operators will be working in, it is possible to mark only that side of the trees. A mark on the base of each tree will allow the forester to check following harvest to see if any seed trees were harvested when they should not have been. Provisions can also be placed in the contract to guarantee retention of undamaged seed trees during the seed-tree cut.

Establishment Treatments

Site preparation including mechanical seedbed preparation and competition control may be necessary to establish the new cohort.

Some species, like eastern cottonwood, require bare mineral soil to germinate. Disturbance of the litter layer using prescribed fire or mechanical means is necessary for the new cohort to establish in these species. Scarification, or the mechanical disturbance of the litter layer, may be accomplished via disking as seen in the photo below. In some situations it may also be possible to sufficiently prepare the seedbed by dispersing skidding throughout the stand to disturb the litter layer.

  • seed_tree6_powell
Figure 3.3.8. Scarification can be done by disking to expose bare mineral soil, as in this ponderosa pine stand in the Pike National Forest of Colorado. Photo Credit: Dave Powell, USDA Forest Service, Bugwood.org

Competition control is as important in naturally regenerating a stand as it is in artificial regeneration. Competition control may be achieved through mechanical means such as disking or mowing, through the use of prescribed fire, or through the application of forest herbicides. These methods are described in greater detail in the site preparation section. One caveat to competition control with the seed-tree regeneration method is that whatever method is selected must not damage the retained seed trees.

Seedling Mortality During the Final Cut

Removing the seed trees during the final cut will result in damaging some trees in the newly established cohort. This is unavoidable, but can be minimized if possible. Aggregating skidding to as small an area as possible will ensure that mortality is minimized. Skidding-based harvest systems in general tend to require that about 10% or less of the harvest area be trafficked (Germain and Munsell 2005).

Maintaining sufficiently high targets for regeneration will also ensure that damage remains within acceptable levels. For example, in longleaf pine 3,000 to 6,000 seedlings per acre are recommended for successful natural regeneration (Franklin 2008). If this target is reached, then damage to 80% of the new cohort, an unreasonably high level under most conditions that will occur during the final harvest, will still result in stocking of 600 to 1,200 trees per acre. This density is sufficient to successfully regenerate the stand under most conditions.

It may become necessary to plant seedlings if damage during the final harvest reduces stocking in the new cohort below acceptable levels across the whole stand or in locations within the stand.

  • seed_tree7_moorhead
Figure 3.3.9. The newly regenerated cohort will be damaged to some extent during the final cut, like these 6-foot-tall lobolly pines. Damage should be minimized where possible, unless a precommercial thin is also prescribed. Photo Credit: David Moorhead, UGA, Bugwood.org

Intermediate Treatments

Naturally regenerated stands often establish at densities far in excess of what is desired for stocking of the mature stand. Precommercial thinning is necessary to maintain a healthy stand that continues to grow at acceptable rates without stagnating due to poor vigor of the individual trees. Precommercial thinning indicates that trees are removed prior to becoming merchantable to improve the growth and vigor of those trees that are retained. Because trees are removed when they are smaller, they are usually left on site. Felling can be by whatever means is most expedient, such as disking or mowing if trees in the new cohort are still small enough. If a precommercial thin is desired at the time of the final cut, it is possible to disperse skidding accordingly to accomplish this goal while the operator is already on the site. Often downed rows will be created, as this is the most efficient way to apply a precommercial thin. This technique is commonly referred to as a corridor thin. Subsequent thins may be commercial.

  • seed_tree8_moorhead
Figure 3.3.10. An overstocked slash pine stand. Young stands with very high densities are common for pines when naturally regenerated. A precommercial thinning before the live crown ratios drop to this point will improve the health of the stand, increasing height growth rates and reducing the risk of insect or disease outbreaks. Photo Credit: David Moorhead, UGA, Bugwood.org
  • seed_tree9_moorhead
Figure 3.3.11. A precommercial thin performed by disking corridors in a young loblolly pine stand. Photo Credit: David Moorhead, UGA, Bugwood.org

Seed-Tree Pros

      • Lower establishment costs (Beaufait et al. 1984).
      • Less labor and machinery required (1984).
      • Provenance of seed is likely locally adapted (1984).
      • Better early root system development (1984).
      • Often less risk of soil movement (1984).
      • Improved aesthetics compared to a clearcut.
      • Improved wildlife habitat compared to a clearcut.
      • Most of value and volume still harvested in the first cut.
      • No certification limits on green-up time or harvest size.

Seed-Tree Cons

      • Less control over spacing and initial stocking (Beaufait et al. 1984).
      • Less control over composition.
      • Produces irregular stands that increases the difficulty of future operations (1984).
      • Risk of seed tree degrading or loss (1984).
      • May not be possible to harvest seed trees on sites with steep slopes or wet soils.
      • No use of genetically improved seedlings (1984).
      • Regeneration delay may cause loss of seedbed or some topsoil and will reduce growth or lengthen the rotation (1984).
      • Longer rotation required.
      • Damaging the new cohort during the final harvest.
      • Precommercial thinning is often required (1984).
      • Less access for fire equipment (1984).
      • Lower economic efficiency versus a clearcut (2 harvest entries, high cost of final harvest).
      • Requires more marking than a clearcut.


  • seed_tree10_hodges
Figure 3.3.12. A mixed loblolly and shortleaf pine stand following a seed-tree cut leaving 8 seed-trees per acre in Crossett, Arkansas. Once the new cohort is established, the seed-trees will be removed. Photo Credit: John Hodges, MSU, Bugwood.org
  • seed_tree11_lockhart
Figure 3.3.13. A mixed black cherry and black birch stand following the seed-tree cut in Pennsylvania. Once the new cohort is established, the seed-trees will be removed. Photo Credit: Brian Lockhart, USDA Forest Service, Bugwood.org
  • seed_tree12_maguire
Figure 3.3.14. A ponderosa pine stand following a seed-tree cut in western Montana. Once the new cohort is established, the seed-trees will be removed. Photo Credit: Doug Maguire, OSU, Bugwood.org
  • seed_tree13_maguire
Figure 3.3.15. A western larch stand following a seed-tree cut in western Montana. Once the new cohort is established, the seed-trees will be removed. Photo Credit: Doug Maguire, OSU, Bugwood.org
  • seed_tree14_mcdonald
Figure 3.3.16. A mixed conifer stand including ponderosa pine, Douglas-fir, sugar pine, California white fir, and incense-cedar following a seed-tree cut at the Challenge Experimental Forest in the Sierra Nevada Mountains of California. A mixture of several light-seeded species with a few heavier seeded species (like sugar pine (Burns and Honkala 1990)) will ensure adequate regeneration. Once the new cohort is established, the seed-trees will be removed. Photo Credit: Philip McDonald, USDA Forest Service, Bugwood.org


Adams, D. L., J. D. Hodges, D. L. Loftis, J. N. Long, R. S. Seymour, and J. A. Helms. 1994. Silviculture Terminology with Appendix of Draft Ecosystem Management Terms. Silviculture Instructors Subgroup of the Silviculture Working Group of the Society of American Foresters. http://oak.snr.missouri.edu/silviculture/silviculture_terminology.htm

Beaufait, W., P. P. Laird, M. Newton, D. M. Smith, C. H. Tubbs, C. A. Wellner, and H. L. Williston. 1984. Silviculture. Pages 413-455 in K. F. Wenger, editor. Forestry Handbook. John Wiley & Sons, New York, NY. ISBN: 0471062278

Burns, R. M. and B. H. Honkala, editors. 1990. Silvics of North America: 1. Conifers. U.S. Department of Agriculture, Forest Service, Washington, DC. http://www.na.fs.fed.us/pubs/silvics_manual/volume_1/pinus/lambertiana.htm

Croker, T. C., Jr. and W. D. Boyer. 1975. Regenerating longleaf pine naturally. USDA Forest Service, Southern Forest Experiment Station, Research Paper SO-105. http://www.treesearch.fs.fed.us/pubs/616

Germain, R., H, and J. F. Munsell. 2005. How much land is needed for the harvest access system on nonindustrial private forestlands dominated by northern hardwoods? Northern Journal of Applied Forestry 22:243-247. http://www.ingentaconnect.com/content/saf/njaf/2005/00000022/00000004/art00004

Franklin, R. M. 2008. Stewardship of Longleaf Pine Forests: A Guide for Landowners. Clemson University Cooperative Extension Service, Clemson, SC. http://www.americaslongleaf.org/media/2558/landowners-guide-to-stewardship-of-llp-2008-_longleaf-alliance_.pdf

Trousdell, K. B. 1950. A method of forecasting annual variations in seed crop for loblolly pine. Journal of Forestry 48:345-348. http://www.ingentaconnect.com/content/saf/jof/1950/00000048/00000005/art00011

Yocom, H. A. 1968. Shortleaf pine seed dispersal. Journal of Forestry 66:422-422. http://www.ingentaconnect.com/content/saf/jof/1968/00000066/00000005/art00016

Hits: 21894