4/27/00 - DoubleTree Hotel, Springfield, OR

Four treatments: light thin (approximately 100-120 TPA), heavy thin (approximately 50-60 TPA), light thin with gaps (approximately .5 acre gaps over 20% of the stand), uncut control

Applied in four blocks (McKenzie RD, Middle Fork RD, and Blue River RD, all in the Willamette National Forest)

Vegetation permanent plots established, .1 hectare in size (approximately .25 acre)

Small quadrats (microplots) used to estimate cover of all vascular plants

Line intercept method used for tall shrubs

Average overstory cover in control plots is 78.9% (SE:1.6%), compared to 79.4% in the immediate post-treatment measurements.

Light thin plots have an average overstory cover of 58.2% (SE:5.2%) in year 3 post-treatment, compared to 50.1% in year one post-treatment measurements.

Gap thin stands averaged 48% (SE:2.2%) in year 3 post-treatment overstory cover, compared to 41% immediately post-treatment.

Heavy thin stands have an average of 34.7% (SE:7.0%) in year 3 post-treatment overstory cover, compared to 31.4% in year one post-treatment

Regression analysis of overstory cover on other explanatory variables such as tree density (conifers only as well as all trees combined) and mean basal area (conifers only and all trees combined) showed significant relationships with each of these.

The best predictor of overstory cover was all-tree mean basal area, R² = 0.83

Understory vegetation was measured by percent cover, and grouped by the following lifeforms:

First, mean percent cover for each understory lifeform category was compared between stands in an ANOVA (ANalysis Of VAriance) to determine if significant treatment effects could be detected.

If a treatment effect did exist, pairwise comparisons of treatment means, using the least significant difference (LSD) method, was used to determine which treatment means differed.

Regression analysis was also performed on understory response with overstory response variables, such as all-tree and conifer mean basal area and tree density, and overstory percent cover.

All significance tests at µ =0.05.

There was no significant treatment effect on bryophyte cover (p=0.1048).

However, bryophyte cover was positively correlated to:

There was a significant treatment response in herb cover (p=.04).

Both the heavy thin and gap thin stands were significantly different than the control stands, but were not significantly different from one another.

Mean herb cover was higher in heavy thin stands at 19.1% (SE:5.7%) and in gap thin stands at 16.7% (SE:3.2%) than in the control stands at 9.9% (SE:2.8%).

Herb cover was negatively correlated to all-tree mean basal area: p=0.02, R²=.67

Herb cover was also significantly negatively correlated to conifer mean basal area: p=0.01, R²=.69

Mean low shrub cover was very similar for all treatments. No significant treatment response was detected.

No significant relationships between low shrub cover and overstory explanatory variables were apparent.

There was a significant treatment response in tall shrub cover (p=0.03).

Mean tall shrub cover from control stands was significantly different (higher) than the other treatment means, but all others treatment means were not significantly different from one another.

Tall shrub cover was significantly positively correlated with:

All-tree mean basal area: p=0.01, R²=0.41

Conifer mean basal area: p=0.01, R²=0.44

All-tree tph: p=0.002, R²=0.51

Conifer tph: p=0.002, R²=0.52

Overstory percent cover: p=0.02, R²=.35

A measure of species richness was assigned to each stand based on understory species recorded.

Both Simpson's and Shannon-Weiner's diversity indices will be calculated for each stand based on individual understory species cover (not completed as of yet).

Both variables were analyzed as described for understory percent cover.

There was a significant treatment response in species richness (p=0.02).

Both light thin and gap thin plots were significantly different (higher) in species richness than control plots.

Species richness was also significantly different (higher) in gap thin plots as compared to heavy this plots.

However, no significant correlations existed between species richness and overstory explanatory variables.

Dense pole plantations do not support a high diversity of bird species because they are structurally simple. Extensive exisitng cover (millions of acres) of dense pole stands is probably is unlikely to occur "naturally" in western Oregon under a natural disturbance regime. We hypothesized that by opening up the canopy and promoting the development of understory layers of vegetation, thinning could increase the diversity of niches available to songbirds, both in the short- and long-term. We tested the effect of 3 different intensities and patterns of thinning: light, heavy, and light with gaps, against unharvested controls. We used standard point count methodology to estimate densities of songbirds for 2 years prior to - and 3 years after application of experimental thinning to 4 blocks of young Douglas-fir stands in the Willamette N.F.

- 7 species that were observed during the pre-treatment phase of the study were not observed during the post-treatment phase: American goldfinch, band-tailed pigeon, Cooper’s hawk, Vaux’s swift, chipping sparrow, lazuli bunting, and ovenbird. However, none of these species was observed consistently across years and blocks before treatment, so their absence afterwards is probably not a treatment effect.

- 14 species were observed during the post-treatment phase but not during the pre-treatment phase: American crow, blue grouse, house wren, mountain quail, northern pygmy owl, northern oriole, olive-sided flycatcher *, red-breasted sapsucker*, red-tailed hawk, Townsend’s solitaire*, white-crowned sparrow, western bluebird, willow flycatcher, and western saw-whet owl. Species with * were observed consistently enough across treatments and years to conclude a positive treatment response.

- Western wood-pewees were observed only once during the pre-treatment phase, but consistently during the post-treatment phase, indicating a positive response to thinning.

- Brown-headed cowbird abundance in one stand increased following thinning.

Cowbirds are brood parasites on the nests of some songbirds, so thinning may have an indirect negative effect on these species if rates of cowbird parasitism increase.

· Species Richness (the average number of species per stand per year) and diversity indices increased in all 3 treatments relative to the unharvested controls following thinning.

· The density of 6 species decreased in one or more treatments relative to controls: golden-crowned kinglet, winter wren, Pacific-slope flycatcher, hermit thrush, Swainson’s thrush, and hermit warbler.

· The density of 4 species (dark-eyed juncos, Hammond’s flycatchers, MacGillivray’s warblers, and western tanagers), and of cavity-nesting birds as a group increased in one or more treatments relative to controls.

· At least 4 species were gained as a result of thinning treatments, and no species appear to have been lost.

· Species richness and diversity of songbird communities increased in response to thinning.

· Some of the species that decreased in density are likely to increase again as canopy closes, and their abundance in treated stands may eventually surpass that in controls, if treated stands develop old-growth like structures more quickly.

· Thinning adjacent to pastures and settlements should be avoided because the rate of brood parasitism by cowbirds may increase and be detrimental to populations of some songbird species.

Ground-dwelling vertebrates are key components of forest ecosystems. They serve as prey for larger vertebrates and play an important role in the dispersal of hypogeous fungi. Understanding how they response to changes in forest conditions is thus critical. Objectives of this study were to determine effects of the three thinning treatments on the relative abundance and diversity of ground-dwelling vertebrates.

Ground-dwelling vertebrates were sampled in the Fall of 1991-92 (pre-treatment) and 1998-99 (post-treatment) using live-traps. Each stand had 100 trapping stations. During a 6-8 consecutive trapping period, 100 Sherman live-traps and 25-50 pitfall traps were employed to record species. Each capture was identified to species and sexed, weighed, and tagged for future recognition.

Eleven small mammal and nine amphibian species were recorded during the four years of this study. Two additional species of voles and two species of shrews were recorded, but identification of these is questionable. The Pacific and fog shrew also were recorded, but positive identification is difficult in the field; these species were recorded as brown shrews. Only eight species of small mammal and one species of amphibian had sufficient captures for further analyses.

1. Are there treatment effects on relative density of species?

A mixed-effects, repeated measures analysis of variance was used to determine treatment effects on capture rates. Statistically significant trends between the pre- and post-treatment periods include:

- increase in relative density of the deer mouse in the two lightly thinned treatments in 1998 (increase in heavy thin treatment but not statistically significant)

- increase in relative density of Ensatina in the two lightly thinned treatments in 1998

- decrease in relative density of Trowbridge’s shrew in the heavy thin in both 1998 and 1999 (same numerical trend was evident for the Pacific & fog shrew aggregate but it was not statistically significant)

- increase in mammalian species’ diversity in the light-thin with gap treatment

Interesting trends in capture rates between pre- and post-treatment periods that were not statistically significant include:

- decrease in western red-backed vole captures in the post-treatment period across all treatment types

- no captures of the northern flying squirrel in the heavy-thin treatments after thinning

- a slight decrease in capture rate of Townsend’s chipmunk in the light-thin with gaps treatments

2. What habitat features are associated with changes in species capture rates?

Regression analysis was used to determine the key habitat features associated with capture rates of a species. Key associations include:

deer mouse - percent herbaceous cover

shrew species - percent moss cover, overstory tree density

Townsend’s chipmunk - density of saplings and percent moss cover

flying squirrel - density of trees >50-cm dbh

Oregon vole - log volume

shrew-mole - density of trees <30-cm dbh

western red-backed vole - log volume

Ensatina - basal area of stumps

3. Is the variability of habitat features within a treatment more or less than that among treatments?

An assumption of the analysis of variance assessment was that all stands of a treatment were in fact representative replicates. The degree of variability in capture rates within a treatment suggested otherwise. Using principal component ordination, variability of habitat features within treatments was found to be slightly less than the variability among treatments. Separate ordinations of habitat features for each species indicated that the most variable stands within a treatment were not necessarily the same among species. Thus, there wasn’t a consistent stand or set of stands that differed from corresponding replicates in terms of stand structure and composition. These results simply go to support the use of the analysis of variance model to analyze treatment effects on capture rates of species.

4. What are the microhabitat characteristics of species? Is there evidence of structural niche segregation?

Using data from only 5 of the 16 stands, principal component ordination indicated extensive overlap in microhabitat-use among small mammal species. Only the two shrew species exhibited some degree of separation in microhabitat-use.

Thinning young Douglas-fir stands had little effect on the ground-dwelling community. Deer mice and Ensatina exhibited a statistically significant numerical increase to the light-thin treatments, at least in 1 of the 2 post-treatment years. Trowbridge’s shrew exhibited a significant decline in the heavy thin treatment in both post-treatment years. However on average, no species was eliminated from a treatment type compared to pre-treatment conditions. An exception was the northern flying squirrel in heavy thin stands, but the variability in capture rates of this species among treatments and all years of the study resulted in a non-significant treatment effect.

Chanterelles are an edible species with significant commercial value

Productivity assessed on a per unit area per year basis

Assessed only control, light thin, and heavy thin treatments

Strip plots and circular plots (8 meter radius)were used; strip plots to measure stand response, circular plots to assess small scale response of chanterelle patches to removal of nearby mycorrhizal host trees

- Commercial grading scheme was not useful

- Dry weights were time consuming and not useful

- Flexible crew scheduling and sampling sequencing were helpful

- Unauthorized harvesting was not a major problem

- High variability among pre-treatment stands and sampling years

- Decrease in productivity following thinning, especially the heavy thin treatment, for both year one and year three post-treatment

- Productivity has not yet shown signs of rebounding in thinned stands

- Showed likelihood of a new chanterelle species

- Identified spatial pattern of distinct individuals within circular plots

- Blind test

- Rainbow and white chanterelle ranked higher than yellow chanterelle

Additional analyses will attempt to separate effects of logging activities from the treatment

Presentation of educational forum options for a series of events in 2001, followed by a small group exercise to help determine content of forums.

1. Now we can evaluate proposed thins on their effect on spotted owl food supply and predict the time anticipated to develop dispersal habitat. We are better prepared to evaluate creation of elk thermal cover and cavity excavator habitat.
2. This study begins to answer the questions we have about whether young stand thinning is beneficial to wildlife and whether it will accelerate late-successional habitat conditions. It begins to illustrate the short-term and long-term trade-offs at the stand scale. This study may assist us in finding appropriate prescriptions to minimize impacts but still achieve overall objectives.
3. Illustrates the challenge of getting USFW (and others) to understand stand dynamics and tree physiography in working with LSR’s. For coastal stands, might as well cut us off at 40-50 years of age (instead of 80) for thinning treatments due to restriction of 20 inches dbh for maximum harvest size.
4. It’s possible to make better management decisions for proposed actions and do better effects analysis.
5. Can’t manage for everything. Need to define clear objectives.
6. Will help determine if land management activities will or will not maintain species viability on a landscape scale.
7. Shows a potential need to restrict collection permits in thinning areas for a number of years to allow fungi to recover.

1. How do legacy features affect development of late-successional habitat in thinned stands?
2. What about thinning effects at the landscape scale?
3. How much, where, what, to thin?
4. Need to know a lot more about thinning effects on everything other than timber: water, air quality, scenery, recreational opportunities, non-timber forest products, wildlife, old-growth dependent species reproduction and dispersal in young stands, etc..
5. Need to know more about effects on survey and manage species.
6. More information on the results of underplanting, (survivability by species, and treatment, as well as growth).
7. Applicability of the results to younger stands (10-20 years old), and to older stands (80-100 years old).
8. More information on basic tree responses: height, growth rates, size and depth of crowns, rate of crown closure, ingrowth, etc.
9. Long-term responses of birds, small mammals, fungi, etc. Keep monitoring.
10. What is the long-term prescription for the stands in the study? Next entry?
11. Continued data collection and monitoring to keep track of changes.
12. We need to know if we can continue to manage vegetation to produce forest products while maintaining healthy populations of fish, wildlife, plants, mushrooms, etc.
13. Should thinning be prescribed in LSR’s and riparian reserves?
14. How the insect community is responding and how they’ll change over time as a reflection of the changing vegetation. Should be done in conjunction with vegetation or wildlife studies.
15. Need to incorporate some training on a collaborative process for how to make the public also a part of the solution.
16. How successful are the underplantings and how fast will they achieve a multi-storied stand? What growth rates can we expect? What about density management, do we precommercially thin?
17. What is the role of natural regen in the thinnings? Can it be managed, and if so, how?
18. What is the response of invertebrates to the different treatements?
19. What is the response of non-native species to thinning? Separate native vs non-native when looking at species richness and diversity. Extend research to bryophytes and herbs.
20. More information on snags: How long do small diameter snags function for wildlife; how long do they function based on method of creation; which species benefit, and at what point in time?
21. When do old-growth dependent species start using these stands?
22. What are the effects on invertebrates?
23. Do birds use these stands primarily for foraging or nesting?
24. Could trees be inoculated with edible mushrooms? Would it hinder snag creation? Will it work on live trees? Could be a potential for non-timber forest products.
25. What is the effect of woody debris and snags on species richness?
26. What is the effect of hardwood pockets vs dispersed hardwoods in the role they provide?
27. More details on strong and weak excavators and between excavators and secondary nesters.
28. What is the long-term effect of thinning on populations of flying squirrel, and what is the implication for managing a known spotted owl home range?

- Small snags also provide foraging habitat

- Snags are reduced in thinning operations through logging operations

- Snags are naturally a part of young stands

Create snags in 2001 across all treatments simultaneously

Target density of 1 snag per acre

Minimum of 12 inoculated and 12 topped in each stand

Create diameter distribution of snags representative of diameter distribution of stand

Create snags in all treatments

Scatter snags, no more than 3 per acre

Use conifers to match composition of stand

Confounding effects with treatments?

Role of insects and disease?

Effectiveness of down logs from tops?

Inoculation methods?

Why 1 per acre?

DAMQUICK - a new method for rapid assessment of stand damage

- >90% of damage accounted for by scarring

- logging systems and silvicultural treatment type did not significantly influence damage levels

- skyline systems had somewhat higher levels

- ground-based logging caused more root damage

- fan-shaped pattern had a lot more damage below landings

- harvester damaged more than twice as many tress as did the forwarder, but forwarder scars were much larger

- damage increased with thinning intensity

damage level depends on definition of minimum scar size

Conducted retrospective studies to examine past damage effects

- once scar heals over decay process is greatly reduced, e.g., 4"-8" wide scar healed in 8 years

- resinous trees heal over more quickly

surveyed mill operators and graders about grade and volume reduction

Then used simulations with Organon to assess future values loss. Either length or diameter deductions are possible. An example of future value loss ($/acre) with a 2 in. diameter deduction on logs harvested 50 years after thinning is show in the table below:

% damage level	damage deduction ($ per acre)
5	58
10.	99
20	189
30	288
40	382

 

To lower stand damage, it is likely that thinning cost will increase. Theseimpacts were simulated for skyline, tractor and cut-to-length thinning with 5% to 20% cycle time increases. For example, with approximately a 10% increase in cycle time (increased time to reduce damage level) logging costs increased by about 9% ($120-$250 per acre depending on the logging system). This analysis indicates that the best economic opportunities (long term economic gains versus increased thinning costs) should be focused on lowering relatively high damage levels to lower levels (e.g. 30-40% lowered to about 10%).

Compared four methods -systematic plots, random plot, transects, blocks

along corridors. All methods had similar results as compared with 100% tally, systematic

plots easiest to install. Damaged trees are concentrated adjacent to corridors for all prescriptions and harvest systems:

- 56% of damage within 15' of the corridor for skyline

- 64% of damage within 15' of the corridor for cut-to-length

- 80% of damage within 15' of the corridor for tractor

- Systematic sample within 15' of corridors

- Comparison with 100% sample on Siuslaw NF showed very close correlation

- Need accurate measurement of harvest area, adequate sample size, and reasonable estimate of how much damage is within 15' of corridors (concentration factor)

- Use tree pads to protect rub trees

- Flag designated skid trails and corridors

- Use straight and adequate width corridors (9' for small yarders, otherwise 12-14')

- Leave low stumps

- Designate trail spacing to keep equipment on trails

- Use parallel yarding pattern whenever possible (as opposed to fan-shaped), but needs to fit terrain

- Change carriage position during lateral yarding

- Match harvester cutting head to the size of the trees

- Choke logs close to the end

- Harvest severely damaged trees

- Intermediate supports reduce skyline lateral excursion

- Time of year and sap flow are important

- Use rub trees as "sacrifice trees" for future snags

 

Project stand and stem growth by treatment

Volume and value through time

Future harvest system analysis and selection

Stand structure measures as a tool to guide achievement of desired habitat attributes

Timing and effect of future entries

Individual tree growth model, distance independent with competition

Input - species and DBH; can add total tree height, crown ratio, 5-year radial growth

Output - stand, stocking, growth and harvest tables; relative density; crown ratio and area; branch diameter

Links to TREEVAL (financial analysis) and VISFOR (visual depiction)

Tracks both stem and stand level characteristics

Stem - crown ratio, taper, branch diameter, bark thickness, volume

Stand - crown closure, vertical and horizontal divesity, volume

Reineke Stand Density Index - stem size and density relationship

Langsaeter Yield-Density - maximize individual tree vs. stand growth

Reviewed density management diagram for Douglas-fir- illustrated use of relative density index to determine zones of competition, full site occupancy and imminent mortality

Use thinning to achieve early financial returns while setting trajectory for future stand structure

Identify future entries to maintain or redirect trajectory

Determine economic feasibility and required harvest systems

Evaluate structural complexity index (Zenner 98)

Heavy thin treatment produces largest diameters and lowest standing volume (cubic feet) at age 80

Thinning at 45 and final harvest at 80 produces much higher revenues than does the control with final harvest at 80; heavy thinning produces slighter higher returns than light thinning

Heavy thin will retain longer crowns

Light thin will move to zone of competition mortality in 20 years, while heavy thin stays in zone of optimal growth for several more decades

What are the tree and stand characterisitcs that matter to obtain desired flora and fauna?

How do we obtain these charactersitics?

What are the implications for harvesting systems and aconomics?

Can you model mixed species and multiple canopy levels?

Will you model the underplanting in the heavy thin?

1) Determine general patterns and trends in stand structure and composition under a range of thinning treatments.

2) Provide general guidelines for future thinning treatments of the stands in the Young Stand Thinning and Diversity Study.

- The ZELIG.PNW (3.0) gap model was used to simulate thinning treatments in a 9-ha representation of one of the actual control plots.

- Treatments consisted of thinning at age 40 to match the heavy thin, light thin, and light-thin with gap treatments, thinning to 4 possible densities at age 60, and thinning to 5 possible densities (including a no-entry treatment) at age 80. Thus, for each of the three age-40 thinning treatments, 20 different thinning combinations were evaluated. The age 40 entry thinned from below; all other entries had an upper diameter limit of 60-cm dbh.

- Variables tracked in each simulation included density of large boles (>100-cm dbh), density of shade-tolerant stems (>40-cm dbh), canopy height diversity, density of snags (>50-cm dbh, >5-m tall), and log mass (>10-cm large-end diameter). Threshold levels of each variable which correspond to old-growth conditions were used to determine the stand age when late-successional conditions developed in a simulation.

General trends include:

The rate at which all live late-successional attributes developed increased with decreasing stand density at age 40 and age 60.

The opposite trend occurred for the dead-wood components. However, retention of 2-4 snags/ha and 5 Mg/ha of logs in the first or second thinning entry was sufficient for threshold levels of these attributes to develop in about the same time as live attributes.

The most limiting live criteria were the rate of development of large boles or vertical diversity. For the age 40 light-thin strategy, vertical diversity was the most limiting across all experimental thinning treatments. For the light thin with gaps and heavy thin initial treatments, vertical diversity developed quickly when thinning heavily at age 60 and development of large boles was the most limiting criterion. For the other age 60 and age 80 thinning strategies, large-bole development was the most limiting criterion.

Rapid development of late-successional attributes does not necessarily result in higher long-term values of attributes. Large-boles developed the quickest for the initial heavy-thin treatment followed by a heavy thinning (thinning to 60 TPH) at age 60. However, this affected future recruitment of stems into the >100-cm size class - by age 160, other thinning strategies which delayed attainment of late-successional conditions by 20-40 yrs had higher densities of large boles. Similar trends for other live attributes, especially vertical diversity, were revealed.

Extracted merchantable volume increased with decreasing stand density at ages 40 and 60.

To more easily evaluate trade-offs among treatments, the extracted volume was divided by the amount of time to satisfy the threshold levels of late-successional attributes (an approach developed by J. Mayo). These results illustrate that the initial heavy-thin treatment with subsequent thinning to <198 TPH at age 60 provided the most extracted volume and fastest attainment of late-successional conditions. Results also illustrate, however, the potential for a range of ‘next’ best treatments which included thinning regimes starting with both the light thin and light thin with gaps.

General guidelines for how stands will respond to thinning treatments were provided by these simulation experiments. In general, heavily thinning the heavy-thin, light-thin, and light-thin w/gaps stands at age 60 without a subsequent thinning entry promotes the most rapid development of late-successional conditions and provides relatively high levels of extracted volume (relative to the other thinning treatments examined in this study). However, deciding which treatment to apply to each of the stands of the Young Stand Thinning and Diversity Study will depend on resource objectives (both short and long-term), specifics of a stand (e.g., susceptibility to windthrow), and economic considerations.

Thanks to Cascade Center for Ecosystem Management and Central Cascades Adaptive Management Area folks for organizing the study and workshop. Also thanks to the personnel at Blue River RD, McKenzie RD and Middle Fork RD personnel for planning and implementing the study. This is just exactly the type of project that was intended in the adaptive management area network. This type of long-term study is priceless; it is absolutely essential to determine the real responses. The value of these kinds of studies just keeps increasing over time. Stick with it. Project includes a good set of integrated studies; you can’t do everything. What’s next? Trying to understand the mechanisms behind why we see the responses that are being reported. Good to hear about the study from a full ecosystem view, not just trees.