1. Experimental gaps and biodiversity responses in the Vermont Forest Ecosystem Management Demonstration Project
Bill Keeton
Graduate student contributors:
Nicholas Dove, Sarah Ford, Heather McKenny, and Kimberly Smith
University of Vermont,
Rubenstein School of Environment and Natural Resources
ABSTRACT
Invtied oral presentation at the 2008 NERC meeting
Biomass development in late-succcessional northern hardwood-hemlock forests: implications for forest carbon sequestration and management
William S. Keeton and Jared Nunery
University of Vermont, Rubenstein School of Environment and Natural Resources, Burlington, VT
Managing the contribution of forest ecosystems to global carbon budgets requires accurate predictions of biomass dynamics in relation to stand development and management. A widely used theoretical model in the northern hardwood region of eastern N. America predicts a peak in biomass after less than two centuries of stand development, followed by declining biomass in stands 200 to 350 years of age, and “steady-state” biomass dynamics in stands > 350 years of age. However, recent empirical studies have found continued basal area accumulations later into stand development than previously predicted. Our study evaluated these competing views, focusing on riparian northern hardwood-conifer forests in the Adirondack Mountains of upstate NY. We sampled 29 sites along 1st and 2nd order stream reaches. Sites were classified as mature forest (9 sites), mature with remnant old-growth trees (5 sites), and old-growth (15 sites). Average age of the largest, dominant trees ranged from approx. 81 to 410 years. At each site forest structure was sampled using 6-10 variable radius plots. We calculated tree biomass based on tree group-specific allometric equations. ANOVA and multiple comparisons were used to analyze categorical data; continuous variables were evaluated using Classification and Regression Trees and linear regression analysis. Aboveground biomass was significantly (p <0.001) different among mature (165 Mg/ha), mature w/remnants (177 Mg/ha), and old-growth (254 Mg/ha) sites. In CART models, basal area was the strongest predictor of dominant tree age, but aboveground biomass was an important secondary variable. Both basal area (r2 = 0.60) and aboveground biomass (r2 = 0.63 were strongly (p < 0.001) and positively correlated with dominant tree age. Biomass approached maximum values in stands with dominant trees approximately 300 to 400 years of age. Our results support the hypothesis that basal area (live and dead) and aboveground biomass can continue to accumulate very late into succession in northern hardwood-conifer forests. Empirical studies suggest there may be more variability in biomass development than predicted by theoretical models. Primary forest systems, especially those prone to partial or intermediate intensity natural disturbances, are likely to have very different biomass dynamics compared to secondary forests developing early peaks in biomass related to initial even-aged development. These differences have important implications for our understanding of both the quantity and temporal dynamics of carbon storage in old-growth forests. Forest management approaches, such as extended rotations and post-harvest structural retention, emphasizing late-successional forest structural objectives consistently yield higher levels of average carbon storage (aboveground forest biomass, dead and alive, plus wood products only) compared to more intensive management approaches. This conclusion is based on our FVS simulation modeling using FIA data from 32 northern hardwood sites distributed throughout the northeastern U.S.

2. Photo credit: Sarah Ford
Rothwald Old-growth Forest, Austrian Alps

3. Vermont Forest Ecosystem Management Demonstration Project

4. Structural Complexity Enhancement (SCE)

5. Gaps are an Element of the Study
Crown release in SCE resulted in clustered harvesting and small gaps (mean opening size = 0.02 ha)
Modified group selection (mean opening size = 0.05 ha)

6. Artificial gaps (“groups”) specifics:
Gap sizes based on mean 0.05 ha (1/8 acre) disturbance scale (from Seymour et al. 2002)
Gap sizes are irregular
Gap shapes are irregular
Light retention within gaps

7. Study Sites Study Areas: Mount Mansfield State Forest
Jericho Research Forest
Paul Smith’s College (FERDA cooperation)
Mature, multi-aged northern hardwoods
History of thinning and selection harvesting
Mid-elevation, moderate productivity

9. FEMDP Research Response Indicators: Growth and yield
Stand structure and dynamics
Herbaceous vegetation
Birds
Small mammals
Amphibians
Fungi
Soil invertebrates
Soil OM and macro-nutrients
Economic tradeoffs and feasibility
Biomass and carbon
Tree Regeneration
© Al Sheldon

10. Leaf Area Index Changes: Pre-Treatment to Post-Treatment
Spatial Variability:
SCE v. GS; not sign.
SCE & GS > STS; P < 0.05
ANOVA:
Fcrit, 0.05 = 2.867
F =
P < 0.001
F tests for variance:
SCE > STS: P = 0.031
GS > STS: P = 0.010
SCE > GS: P = 0.296
Keeton For. Ecol. and Mgt.

11. Coarse Woody Debris Enhancement

12. Red-backed Salamander Response Based on Occupancy Modeling
Plot of mean abundance of salamanders as a function of standardized covariates. You can see how important it is to have lots of downed logs in later stages of decay for salamander habitat. You also get a sense of the small effect relative density of overstory trees had on salamander abundance.
McKenny, Keeton, and Donovan. 2006

13. Response of Late-successional Understory Plant Species
Richness:*
p = 0.012
SCE > GS
Shannon Index:*
p = 0.009
SCE > CON
* Following Hill’s (1973) series of diversity Indices
Smith, Keeton, Twery, and Tobi CJFR

14. Locally Extirpated Species
ANOVA:
p = 0.07

15. Fungal Responses; Aboveground Sporocarps
Dove and Keeton Fungal Ecology

16. Fungi Responses: Classification and Regression Tree
Initial formula included 7 structural variables
Dove and Keeton Fungal Ecology

17. Biomass and carbon in downed logs 10 years post-harvest
Mg/ha
Treatment Type

18. Closing Thoughts
Silvicultural gaps promote some elements of late-successional biodiversity, depending on within gap structure
Spatial configuration w/closed canopy patches also important
Manage for temporal and spatial variability
There is no “one-size-fits all” approach; mix it up!
Adapt, learn from unanticipated results

19. Acknowledgements Vermont Monitoring Cooperative
U.S. National Science Foundation
Northeastern States Research Cooperative
USDA McIntire-Stennis Forest Research Program
USDA National Research Initiative