2012 Consolidated WBP Direct Seeding Trials

Project: Consolidated WBP Direct Seeding Trials

Attachment:  Various_Roberts_2012_APPENDIX

Agency/Forest or Park/District: Montana State University

Project coordinator: Dave Roberts

Contact:  David W. Roberts, Professor and Head, Department of Ecology, Montana State University, Bozeman, MT 59717-3460, office 406-994-4548, FAX 406-994-3190, email droberts@montana.edu

Clay DeMastus, 200 Michael Grove Ave.,  Bozeman, MT 59718; cdemastus@hotmail.com.


Valerie Walker, and John Waverek, Lolo NF; Sidnee Dittman, St. Joe NF;  Dan Reinhart, Yellowstone National Park;  Bev Yelczyn, Clearwater NF; Nancy Lankford, Mt. Hood National Forest;  Richard L. Roberson, Caribou-Targhee NF, Grand Targhee Ski Area;  Melissa Jenkins, Flathead National Forest;  Jodie Canfield, Mark Novak, Gallatin National Forest;  Cda Nursery; Cathy Cripps, Montana State University.

In addition to these direct cooperators, several scientists have been involved in developing this project and will continue to be advisors/consultants for changes or additional treatments. (This includes John Schwandt, Holly Kearns, Diana Tomback, Ward McCaughey, and others.)

Source of funding /amount

FHP: $15,000

Supplemental funding: $15,000 from MSU, FS and Nursery

Dates of restoration efforts

Fall 2012


  • Investigate what treatments (if any) are most effective at increasing germination of directly sown seeds.
  • Investigate what treatments (if any) are most effective at increasing survival of germinants from directly sown seeds.
  • Investigate if caging affects germination of directly sown seeds and survival of germinants from directly sown seeds.
  • Compare long term survival of directly sown seeds versus long term survival of outplanted nursery grown seedlings.
  • Compare germination and survival rate of directly sown seeds in seed caches versus non-cached or individually directly sown seeds.

Acres/ha treated  N/A


Six sites throughout the northern Rocky Mountains in the states of Montana and Idaho were chosen to carry out the field experiments described in this chapter.  These sites were: Fairy Lake on the Gallatin National Forest near Bozeman, MT, Pioneer Mountain within The Yellowstone Club Ski Area near Big Sky, MT, Toboggan Ridge on the Clearwater National Forest in Idaho, Thompson Peak on the Lolo National Forest near Plains, MT, Ulm Peak on the Idaho-Montana state line west of Thompson Falls, MT, and Gold Pass on Idaho-Montana state line west of St. Regis, MT. These sites were chosen for their whitebark pine habitat and accessibility for ease in set-up and data collection.

In this portion of the experiment, 800 seeds collected from local seed sources were used at each site.  The seeds were planted in a complete randomized block design with five replicates of eight treatments with 20 seeds per treatment.  Each of the 20 seeds in each treatment was planted 3.25″ apart from one another at a depth of 1.5″ in bare mineral soil.  The treatments included; 1) warm stratification (strat), 2) scarification (scar), 3) both the stratification and scarification together (scar + strat), and 4) a control.

These four treatments were installed with and without a 1/2 ″ mesh wire cage in each replicate to make up the eight blocks and treatments. Figure 1 is an example that shows the layout of one site broken down by blocks and treatments.  This consisted of eight rows of eight treatments spaced approximately three feet apart for ease in data collection with 20 seeds in each row.  Two HOBO® data loggers were also installed at each site to record ground surface temperatures as well as soil temperatures 1.5″ below the surface (depth of sown seed).  These were placed in selected caged treatments in various locations throughout the site.


Single Site 5 Blocks Single Block
(20 Seeds Per Treatment)
Block 1
Block 2 Stratification
Scarification + Stratification
Block 3 Stratification:Cage
Block 4 Scarification:Cage
Scarification + Stratification:Cage
Block 5 Scarification

Figure 1:  Schematic layout of a site.

Nursery Seedlings and Seed Caches

Two year old nursery seedlings grown from local seed sources were planted at each site spread throughout the areas encompassing the five blocks.  Thirty-four nursery seedlings were planted at the Gold Pass and Thompson Peak sites while 100 were planted at the remaining sites.

At the Fairy Lake, Thompson Peak, Ulm Peak, and Gold Pass sites, caches of three untreated seeds were planted directly next to 32 of the nursery grown seedlings.  An untreated seed cache of four was planted next to one of the nursery grown seedlings for a total of 100 seeds in 33 caches.  At the Pioneer Mountain and Toboggan Ridge sites, untreated seed caches of three were planted next to 50 of the nursery grown seedlings and warm stratified seed caches of three were planted next to the remaining 50 nursery grown seedlings for a total of 300 seeds in 100 caches.


This study’s initial intent was to analyze the results of direct seeding tests performed in six whitebark pine habitats throughout the northern Rocky Mountains.  One of these six sites (Fairy Lake) had very low germination rates and even poorer survival rates.  This particular site is unique because it has both limber pine (Pinus flexilus) and whitebark pine growing within the same general area.  In Montana, this is a rarity as there are only few sites that posses this same characteristic (Pfister et al. 1977).  Limber pine is phenotypically similar to whitebark pine.  The best visual discerning feature that separates the two species is the difference in their unique cones.  Whitebark and limber pine use mast seeding to improve their fitness.  Some years there are bountiful amounts of cones and other years it is difficult to find even one cone on a particular tree.  The year the Fairy Lake site was installed (2009) was not a mast year.  Because of this, the difference between the two species at this site was very difficult to discern.  The site was thought to be installed in whitebark pine habitat because there where known whitebark pine plus trees that where used in past cone and scion collections by local Forest Service crews close by and were marked with identification tags.  However, throughout the course of the study, it became apparent in mast years (2010 and 2011) that the site had been installed in a zone transitioning from limber habitat to whitebark habitat.  It is thought that this along with the fact that it was a dry, rocky and more harsh site could have been the reason such poor germination and survival occurred here.  Because of this, the decision was made to not include the data from the Fairy Lake site in the final analysis.


Germination results are presented below in tables 1 and 2.

Table 1: Excerpt of the summary of the final model of germination.


Estimate     Std. Error    t value     Pr(>|t|)

(Intercept)                                 -9.592e-01    4.308e-01   -2.226    0.027650

sitepioneer                                 -4.254e-01    6.200e-01   -0.686    0.493810

sitethompson                              3.483e-01    6.050e-01     0.576    0.565746

sitetoboggan                               1.311e+00   5.870e-01     2.233    0.027199

siteulm                                        2.014e-01    6.043e-01     0.333    0.739397

seedscar                                     -6.190e-01    3.335e-01    -1.856    0.065591

seedscar+strat                            4.171e-01    3.203e-01   1.302    0.195087

seedstrat                                      1.299e+00   3.339e-01   3.891    0.000156

sitepioneer:seedscar                 1.526e+00      4.832e-01   3.158    0.001960

sitethompson:seedscar            -2.116e-01       4.922e-01  -0.430    0.667936

sitetoboggan:seedscar              4.911e-01       4.612e-01   1.065    0.288851

siteulm:seedscar                      -7.624e-01       5.488e-01  -1.389    0.167047

sitepioneer:seedscar+strat       -1.038e-01       4.768e-01  -0.218    0.827923

sitethompson:seedscar+strat   5.249e-01       4.615e-01   1.137    0.257411

sitetoboggan:seedscar+strat    -7.191e-01       4.534e-01  -1.586    0.115078

siteulm:seedscar+strat              1.809e-01       4.625e-01   0.391    0.696288

sitepioneer:seedstrat                -2.715e-01       4.831e-01  -0.562    0.575062

sitethompson:seedstrat            -5.780e-01       4.700e-01  -1.230    0.220905

sitetoboggan:seedstrat            -1.470e+00   4.618e-01  -3.183    0.001807

siteulm:seedstrat                      -3.002e-01      4.724e-01  -0.636    0.526156


Null deviance: 928.68  on 199  degrees of freedom

Residual deviance: 339.13  on 135  degrees of freedom

Table 2:  Estimated germination percent and 80% confidence intervals at each site by treatment.

Germination Estimates
Site Treatment Estimated Germination (%) 80% CI (%)
Gold Pass Control 27.8 17.4 to 41.0
Gold Pass Scar 17.1 10.0 to 27.7
Gold Pass Scar + Strat 36.8 24.4 to 51.2
Gold Pass Strat 58.4 43.5 to 71.9
Pioneer Mountain Control 20.0 11.9 to 31.7
Pioneer Mountain Scar 38.3 25.5 to 52.9
Pioneer Mountain Scar + Strat 22.5 13.7 to 34.6
Pioneer Mountain Strat 41.5 27.9 to 55.9
Thompson Peak Control 35.2 23.2 to 49.4
Thompson Peak Scar 19.1 11.3 to 30.6
Thompson Peak Scar + Strat 58.2 43.8 to 71.4
Thompson Peak Strat 52.8 38.3 to 66.7
Toboggan Ridge Control 58.7 45.0 to 71.2
Toboggan Ridge Scar 55.6 41.9 to 68.5
Toboggan Ridge Scar + Strat 51.2 37.7 to 64.6
Toboggan Ridge Strat 54.5 40.9 to 67.5
Ulm Peak Control 31.9 20.7 to 45.7
Ulm Peak Scar 10.5 5.8 to 18.5
Ulm Peak Scar + Strat 46.0 32.4 to 60.2
Ulm Peak Strat 56.0 41.7 to 69.3

As can be seen from table 2, stratification generally results in the best germination results, except at Thompson Peak where scarification+stratification performed best, and at Toboggan Ridge where treatments did not differ.


Survival results are presented in Tables 3 and 4.

Table 3: Excerpt of the summary of the final model of survival.


Estimate    Std. Error    t value     Pr(>|t|)

(Intercept)                                 -0.43262        0.57004   -0.759     0.449178

sitepioneer                                  1.32350        0.82193    1.610     0.109614

sitethompson                              1.68553        0.71830    2.347     0.020360

sitetoboggan                               2.37399        0.74875    3.171     0.001872

siteulm                                       -1.48856        0.74456   -1.999     0.047531

seedscar                                      0.06709        0.45638     0.147     0.883346

seedscar+strat                           -0.18488        0.39546    -0.467     0.640879

seedstrat                                    -0.57778        0.36279    -1.593     0.113519

sitepioneer:seedscar                  -0.04934        0.63264     -0.078    0.937942

sitethompson:seedscar              -0.72066        0.69032     -1.044    0.298319

sitetoboggan:seedscar               -0.61050        0.68719     -0.888    0.375858

siteulm:seedscar                        -0.30251        0.90995     -0.332    0.740053

sitepioneer:seedscar+strat         -1.02198        0.60630     -1.686    0.094114

sitethompson:seedscar+strat     -0.19623        0.55727     -0.352    0.725271

sitetoboggan:seedscar+strat       0.20137           0.68780      0.293    0.770133

siteulm:seedscar+strat             1.46592           0.61491      2.384    0.018479

sitepioneer:seedstrat                0.07272           0.55811      0.130    0.896522

sitethompson:seedstrat            0.69073           0.54866      1.259    0.210161

sitetoboggan:seedstrat            0.68328           0.67702      1.009    0.314612

siteulm:seedstrat                     2.02245           0.58307      3.469    0.000697


Table 4:  Estimated survival percent and 80% confidence intervals at each site by treatment.

Survival Estimates
Site Treatment Survival (%) 80% CI (%)
Gold Pass Control 39.4 26.0 to 54.5
Gold Pass Scar 41 27.5 to 56.0
Gold Pass Scar + Strat 35 23.1 to 49.2
Gold Pass Strat 26.7 17.5 to 38.5
Pioneer Mountain Control 70.9 56.4 to 82.1
Pioneer Mountain Scar 71.3 57.8 to 81.8
Pioneer Mountain Scar + Strat 42.2 32.5 to 52.5
Pioneer Mountain Strat 59.5 45.5 to 72.2
Thompson Peak Control 77.8 68.7 to 84.8
Thompson Peak Scar 64.6 51.2 to 76.0
Thompson Peak Scar + Strat 70.5 61.0 to 78.5
Thompson Peak Strat 79.7 71.5 to 85.9
Toboggan Ridge Control 87.5 80.6 to 92.1
Toboggan Ridge Scar 80.2 71.1 to 86.9
Toboggan Ridge Scar + Strat 87.6 80.3 to 92.5
Toboggan Ridge Strat 88.6 81.5 to 93.2
Ulm Peak Control 12.8 8.1 to 19.7
Ulm Peak Scar 10.4 4.9 to 20.8
Ulm Peak Scar + Strat 34.5 26.4 to 43.7
Ulm Peak Strat 38.3 30.0 to 47.4

Survival varied significantly by site but stratification was helpful at all but one site.

Combined Germination and Survival

Previous results detailed germination and survival of germinants separately, but clearly it is the product of both that is significant to future whitebark pine.  Combined germination and survival statistics are presented in Tables 5 and 6.

Table 5:  Excerpt of the summary of the final model of combined germination and survival.


Estimate    Std. Error    t value      Pr(>|t|)

(Intercept)                                 -2.14591       0.53677    -3.998     0.000103

sitepioneer                                  0.37338       0.71148      0.525     0.600554

sitethompson                              1.43538       0.67178      2.137     0.034351

sitetoboggan                               2.28682       0.64660      3.537     0.000549

siteulm                                       -0.96513       0.72999     -1.322     0.188271

seedscar                                     -0.50209       0.35262     -1.424     0.156689

seedscar+strat                            0.19741       0.32645      0.605     0.546337

seedstrat                                     0.44583       0.31686      1.407     0.161621

sitepioneer:seedscar                1.25092       0.49292      2.538     0.012243

sitethompson:seedscar            -0.57429      0.54868     -1.047     0.297048

sitetoboggan:seedscar             0.20929       0.46074      0.454      0.650344

siteulm:seedscar                      -0.73468      0.88739     -0.828     0.409123

sitepioneer:seedscar+strat       -0.53308      0.50279     -1.060     0.290853

sitethompson:seedscar+strat               0.19249       0.46788      0.411     0.681407

sitetoboggan:seedscar+strat    -0.51369      0.44140     -1.164     0.246482

siteulm:seedscar+strat             1.09962       0.57993      1.896     0.059991 .

sitepioneer:seedstrat                0.13055       0.46984      0.278     0.781528

sitethompson:seedstrat            0.07096       0.45988      0.154     0.877592

sitetoboggan:seedstrat            -0.60093      0.43202     -1.391     0.166428

siteulm:seedstrat                     1.30631       0.56441      2.314     0.022086


Table 6:  Estimated combined germination and survival percent and 80% confidence intervals at each site by treatment

Combined Germination and Survival Estimates
Site Treatment Estimated Germ. + Surv. (%) 80% CI (%)
Gold Pass Control 10.5 5.9 to 18.0
Gold Pass Scar 6.6 3.6 to 11.9
Gold Pass Scar + Strat 12.5 7.1 to 21.0
Gold Pass Strat 15.5 9.0 to 25.3
Pioneer Mountain Control 14.5 8.6 to 23.5
Pioneer Mountain Scar 26.4 16.9 to 38.8
Pioneer Mountain Scar + Strat 10.8 6.2 to 18.3
Pioneer Mountain Strat 23.2 14.6 to 34.9
Thompson Peak Control 33.0 22.5 to 45.3
Thompson Peak Scar 14.3 8.5 to 23.2
Thompson Peak Scar + Strat 42.1 32.8 to 51.9
Thompson Peak Strat 45.2 32.9 to 58.0
Toboggan Ridge Control 53.5 41.6 to 65.0
Toboggan Ridge Scar 46.2 34.7 to 58.1
Toboggan Ridge Scar + Strat 45.6 34.2 to 57.6
Toboggan Ridge Strat 49.7 37.9 to 61.4
Ulm Peak Control 4.3 2.7 to 6.5
Ulm Peak Scar 1.3 0.5 to 3.2
Ulm Peak Scar + Strat 14.0 11.0 to 17.8
Ulm Peak Strat 20.4 13.8 to 29.1




Cage Effects 


Table 7: Caged versus non-caged estimated germination rates comparison by treatment and site.


Site Treatment Non-Caged Germ. (%) Caged Germ. (%)
Gold Pass Control 27.8 29.6
Gold Pass Scar 17.1 18.4
Gold Pass Scar + Strat 36.8 38.9
Gold Pass Strat 58.4 60.6
Pioneer Mountain Control 20.0 36.3
Pioneer Mountain Scar 38.3 58.5
Pioneer Mountain Scar + Strat 22.5 39.7
Pioneer Mountain Strat 41.2 61.4
Thompson Peak Control 35.2 20.6
Thompson Peak Scar 19.1 10.2
Thompson Peak Scar + Strat 58.2 40.0
Thompson Peak Strat 52.8 65.2
Toboggan Ridge Control 58.7 50.0
Toboggan Ridge Scar 55.6 46.8
Toboggan Ridge Scar + Strat 51.2 42.5
Toboggan Ridge Strat 54.5 45.7
Ulm Peak Control 31.9 35.6
Ulm Peak Scar 10.5 12.2
Ulm Peak Scar + Strat 46.0 50.1
Ulm Peak Strat 56.0 60.0




Table 8.  Cage and Non-Caged Survival Estimates
Site Treatment Non-Caged Surv. (%) Caged Surv.(%)
Gold Pass Control 39.4 61.2
Gold Pass Scar 41.0 63.5
Gold Pass Scar + Strat 35.0 57.5
Gold Pass Strat 26.7 47.7
Pioneer Mountain Control 70.9 84.8
Pioneer Mountain Scar 71.3 85.1
Pioneer Mountain Scar + Strat 42.2 62.6
Pioneer Mountain Strat 59.5 77.2
Thompson Peak Control 77.8 74.9
Thompson Peak Scar 64.6 60.9
Thompson Peak Scar + Strat 70.5 67.1
Thompson Peak Strat 79.7 77.0
Toboggan Ridge Control 87.5 94.8
Toboggan Ridge Scar 80.2 91.4
Toboggan Ridge Scar + Strat 87.6 94.9
Toboggan Ridge Strat 88.6 95.3
Ulm Peak Control 12.8 25.6
Ulm Peak Scar 10.4 21.4
Ulm Peak Scar + Strat 34.5 55.4
Ulm Peak Strat 38.3 59.4



Future monitoring of test sites will depend on the availability of funding; no plans exist at present to undertake such monitoring.


Whitebark pine seed treatments of warm stratification and warm stratification and scarification together were found to have significantly increased germination under natural field settings compared to the scarification treatment at two sites.  Warm stratified seed was found to have significantly higher germination rates than scarified seed and untreated control seed at one of the other three sites.  As a whole, warm stratified seed had the highest germination rates as estimates did not drop lower than 41% at any of the sites.  The other three treatments had germination rates below this mark at multiple sites.

Survival rates of whitebark pine germinants from directly sown seed were similar between treatments with a few exceptions.  The control and scarification treatments were found to have significantly higher survival rates than seed undergoing both scarification and warm stratification at one of the five sites.  Warm stratified seed and seed undergoing both stratification and scarification were found to have significantly higher survival rates than the untreated control and scarified seed at one of the remaining four sites.

Warm stratified seed and seed undergoing both warm stratification and scarification were found to have significantly higher combined germination and survival rates than scarified seed at one of the five sites and significantly higher combined germination and survival rates than untreated control seed and scarified seed at the one of the remaining four sites.  Warm stratified seed had the highest estimated combined germination and survival rates at three of the five sites.  This same treatment had the second highest estimated combined germination and survival rates at the remaining two sites and was surpassed by the treatment with the highest estimated rates by only three percent at both of these two sites.

Caging directly sown seed was shown to have a positive effect on germination and survival that is dependent on both site and block.  Caging was also shown to have a positive effect on survival that is dependent on seed treatment.  It is thought that this caging effect on survival could be due to an unexpected cooling effect brought about by the shade provided by the cage.  The shade provided by the cage is minimal but possibly significant enough to decrease temperatures, which results in a decrease in evapotranspiration, ultimately resulting in an increase in survival.

Outplanted nursery grown seedlings had significantly higher survival rates than the combined germination and survival rates of seeds sown in caches after two and three years of observation.  Ideally the data concerning this aspect of the experiment will continue to be recorded on a yearly basis in the future to monitor long term survival.

Cache seed germination rates were lower than seeds planted alone (non-cached) at all sites except one, Thompson Peak.  Cache seed survival rates and non-cache seed survival rates were very similar at all five sites as no significant difference was found.

Soil surface temperatures had higher volatility than subsurface temperatures and winter temperatures stayed very close to freezing once there was snow cover. The number of extreme temperature events at a site is estimated to increase survival of germinants from directly sown seeds.  However, this is probably not the cause of increased survival and results such as this are likely an indicator that one or many other important variables affecting survival are not accounted for in this study.