Elsevier

Forest Ecology and Management

Volume 432, 15 January 2019, Pages 1053-1063
Forest Ecology and Management

Short-term, spatial regeneration patterns following expanding group shelterwood harvests and prescribed fire in the Central Hardwood Region

https://doi.org/10.1016/j.foreco.2018.10.043Get rights and content

Highlights:

  • Prescribed fire may be used in group shelterwoods to regenerate oak on mesic sites.

  • The highest density of competitive oak regeneration was generally outside of harvested groups.

  • Oak regeneration patterns were affected by cardinal direction from group center.

  • Prescribed fire did not affect competitive oak seedlings, but did reduce major competitors.

Abstract

Throughout eastern North America, oaks (Quercus) are a foundational tree species, but are regenerating poorly, particularly on mesic sites. This regeneration failure has spurred development of new management practices that create heterogeneous regeneration conditions that better match oak’s response to disturbances such as surface fire and windthrow. Expanding group shelterwood systems are designed to produce diverse regeneration conditions and have a high edge-to-forest interior ratio, where intermediate light levels may be beneficial for oak regeneration. We present data on early regeneration patterns from a large-scale experiment designed to assess the combined effects of these silvicultural systems and prescribed fire on oak regeneration, ecosystem resilience, and spatial and compositional heterogeneity in the Central Hardwood Region. Using transect-based surveys, we investigated the spatial patterns of woody regeneration within and outside of burned and unburned 2- or 3-stage group shelterwoods in factorial replicates at two different sites. Two years following the initial harvest, the south-facing site had substantial competitive oak regeneration just outside of the harvested groups on the northern, eastern, and western sides, but the east-facing site did not. On a stand level, tulip poplar (Liriodendron tulipifera) and sassafras (Sassafras albidum) regeneration increased in both sites, oak increased in the south-facing site, hickory (Carya) increased in the east-facing site, and maple (Acer) was relatively unaffected by the treatments. Competitive oak regeneration in the forest matrix just outside of the harvested groups in the south-facing site holds promise to regenerate a stand with a substantial oak component and high overall diversity, given the shelterwood groups will be expanded outward in successive entries and burned repeatedly over time.

Introduction

The current structure and composition of many North American forests reflects decades of fire suppression and production-focused management practices that homogenized stands to meet rigid compositional and structural targets, resulting in stands with diminished resilience and ecological memory (Drever et al., 2006, Guyette et al., 2002, Long, 2009, Puettmann et al., 2009, Webster et al., 2018). Faced with future climatic uncertainties and disease outbreaks, restoring forest diversity and resiliency while still meeting production objectives is of high concern (Mori et al., 2013, Puettmann, 2011). Managers and ecologists increasingly recognize that restoring resilience requires reassessing traditional harvesting methods to better align with an ecosystem’s natural disturbance regime (Drever et al., 2006, Puettmann and Ammer, 2007). While natural disturbance-based approaches are theoretically promising (Franklin et al., 2007, Long, 2009), large-scale research on these approaches is sparse and attempts to apply these systems are not always successful (Fahey et al., 2018, Kern et al., 2017, Palik et al., 2002).

In eastern North America, many hardwood forests are dominated by mature oak (Quercus) and hickory (Carya) overstory, but have very little oak regeneration (Abrams, 2003, Aldrich et al., 2005). Loss of oak as a dominant canopy species will cause substantial changes in resource availability and cascading trophic effects throughout deciduous forests of eastern North America (McShea et al., 2007, Smith, 2006). Factors implicated in oak regeneration failure include: fire suppression; reduction of small canopy gaps; invasive species; and increased herbivory (Guyette et al., 2002, Nowacki and Abrams, 2008). Regardless of the contributing factors, ultimately, contemporary management practices frequently do not match oak’s adaptations to disturbance and, therefore, fail to provide conditions necessary for adequate regeneration (Arthur et al., 2012, Dey, 2002, Jenkins and Parker, 1998).

Many eastern oak species are considered intermediately shade tolerant and fire-adapted (e.g., resprouting ability, thick bark at maturity, and hypogeal germination); however, traditional management often does not align with these traits (Arthur et al., 2012, Dey, 2002, Johnson et al., 2009). In unharvested or lightly harvested stands (e.g., single-tree selection), more mesic, shade tolerant species such as maple (Acer) and beech (Fagus) typically dominate the regeneration layer, whereas complete or heavy overstory removal (e.g., clearcuts or large group selection openings) shifts composition to faster growing, early successional species such as tulip poplar (Liriodendron tulipifera) and sassafras (Sassafras albidum; Dey, 2002, Swaim et al., 2016). Furthermore, after decades of fire suppression, and corresponding increase in understory density, single prescribed fires do little to modify regeneration patterns (Brose et al., 2013, Alexander et al., 2008, Dey and Fan, 2009, Hutchinson et al., 2012). Generally, oak regeneration is most successful when silvicultural methods emulating partial or patchy stand mortality are used in concert with prescribed fire (Brose et al., 1999, Hutchinson et al., 2012, Kern et al., 2017).

There is some evidence that oaks may regenerate well on the edge of and just outside of gaps where light levels are intermediate; however, most studies solely focus on regeneration within harvest gaps (Lhotka and Stringer, 2013, Schmidt and Klumpp, 2005, Schulte et al., 2011). Expanding group shelterwoods (patterned after a Bavarian or Bayerischer Femelschlag; Puettmann et al., 2009), remove small percentages of a stand in a series of expanding, small- to medium-sized canopy openings similar to those caused by wind or tree senescence, and create stands with high structural, age class, and species diversity (Seymour, 2005). While this silvicultural regeneration system has been used in North American coniferous and mixedwood systems (Arseneault et al., 2011, Raymond et al., 2009), it remains largely untested in hardwood systems. Expanding group shelterwoods maintain a high edge-to-forest interior ratio that might promote advanced oak regeneration in the forest matrix directly outside of harvest groups in the area slated for subsequent harvests (Arseneault et al., 2011, Lhotka and Stringer, 2013).

While altered regeneration along “high-contrast” ecological edges (e.g., roads, paths, pasture boundary) is well documented; there is little empirical information describing “lower-contrast” edges, such as intra-stand harvested gaps altering regeneration in the surrounding forest matrix (Arseneault et al., 2011, Lhotka and Stringer, 2013, Matlack, 1993, Schmidt and Klumpp, 2005). The forest understory is typically a light-limited system, and harvest boundaries dramatically alter this resource along a spatial gradient extending from within the harvest opening into the adjacent forest matrix (Lhotka and Stringer, 2013, Voicu and Comeau, 2006). The light that filters through a harvest gap into the adjacent forest matrix is affected by gap size, percent of overstory removed within the gap, and edge orientation (N, E, S, W; Matlack, 1993).

Beginning in 2014, a landscape-scale, temporally replicated expanding group shelterwood and prescribed fire experiment was initiated in southern Indiana to promote oak regeneration and increase stand-level heterogeneity. Past research and oak autecology suggest that shelterwood groups will produce areas with intermediate light levels required for oak regeneration and the concurrent use of prescribed fire should further reduce shade-tolerant, but fire-sensitive competitors; however, these two treatments have never been tested together in an expanding group shelterwood system (Brose et al., 1999, Hutchinson et al., 2012, Loftis, 1990). We present results from an exploratory study designed to assess early spatial regeneration patterns in the first two replicates of this study, each of which contained four factorial treatments: 2- and 3-stage expanding group shelterwoods, with and without prescribed fire. Specifically, we investigated how oak, hickory, maple, sassafras, and tulip poplar regeneration patterns were affected by group shelterwood and prescribed fire treatments and orientation (N, E, S, W) within and outside of initial shelterwood gaps. Additionally, we assessed how competitive oak regeneration was affected by the presence of specific non-oak seedlings, basal area, and canopy cover. We concentrated our analyses on the presence of established seedlings, rather than overall abundance, to better predict future composition and available advanced regeneration for subsequent gap expansions (Iverson et al., 2008).

Section snippets

Study site

This research was conducted at Naval Support Activity (NSA) Crane in Martin County, Indiana. Most of the 210 km2 of forested land on NSA Crane is secondary growth originating from the 1850s–1930s; has a similar history to many forests in the Central Hardwood Region; is managed predominantly for oak and hickory; and is relatively unaffected by base operations. Currently, harvests at NSA Crane total approximately 525–600 ha per year (3 MMBF per year), which accounts for about 40% of annual growth

Replicate one

The proportion of quadrats stocked with an established oak, sassafras, and tulip poplar seedling increased two years following burn and shelterwood harvests in both the stand-level and gap-focused surveys (Fig. 2). The proportion of established hickory and maple seedlings did not change from the pre-treatment survey to either of the post-treatment surveys (Fig. 2).

Established oak regeneration >30 cm tall displayed a significant interaction between position within the stand and direction, which

Discussion

Natural regeneration patterns can be difficult to quantify and predict following ecological-based silvicultural treatments because these treatments are designed to regenerate compositionally and structurally diverse stands (Dey, 2014, Kern et al., 2017, Webster et al., 2018). In these systems regeneration patterns are influenced by a suite of interacting factors including resource, seed tree, and germination substrate availability; competition with advanced regeneration; herbivory; and seed

Declarations of interests

None

Acknowledgments:

Trent Osmon, Brady Miller, and Rhett Steele of NWSC Crane provided logistical and on the ground support, without which this project would not have been possible. Rob Swihart and Ken Kellner provided study design and analysis assistance and valuable feedback on the manuscript. Laura Estrada, Ethan Belair, David Ralston, Ryan Bartlett, Julia Buchanan-Schwanke, James McGraw, Matt Moore, Landon Neuman, and Ben Taylor, helped with data collection in the field. Mike Steele and two anonymous reviewers

Funding

This work was supported by the Department of the Navy (Cooperative Agreement number N62470-14-2-9001), the Department of Forestry and Natural Resources at Purdue University, and McIntire-Stennis Cooperative Forestry Research Program (project number IND011557MS).

References (62)

  • M.F. Voicu et al.

    Microclimatic and spruce growth gradients adjacent to young aspen stands

    For. Ecol. Manage.

    (2006)
  • M.D. Abrams

    Where has all the white oak gone?

    BioScience

    (2003)
  • M.D. Abrams et al.

    The impacts of mast year and prescribed fires on tree regeneration in oak forests at the Mohonk Preserve, Southeastern New York, USA

    Nat. Areas J.

    (2013)
  • M.D. Abrams et al.

    Long-term seedling height growth and compositional changes following logging and wildfire in a Central Pennsylvania oak forest

    Castanea

    (2013)
  • P.R. Aldrich et al.

    Confirmation of oak recruitment failure in Indiana old-growth forest: 75 years of data

    For. Sci.

    (2005)
  • M.A. Arthur et al.

    Refining the oak-fire hypothesis for management of oak-dominated forests of the Eastern United States

    J. For.

    (2012)
  • T.A. Barnes et al.

    Prescribed fire effects on advanced regeneration in mixed stands

    South. J. Appl. For.

    (1998)
  • D. Bates et al.

    Fitting linear mixed-effects models using lme4

    J. Stat. Softw.

    (2009)
  • P.H. Brose et al.

    A meta-analysis of the fire-oak hypothesis: does prescribed burning promote oak reproduction in Eastern North America?

    For. Sci.

    (2013)
  • Brose, P.H., Gottschalk, K.W., Horsley, S.B., Knopp, P.D., Kochenderfer, J.N., McGuiness, B.J., Miller, G.W., Ristau,...
  • P.H. Brose et al.

    A comparison of the survival and development of the seedlings of four upland oak species grown in four different understory light environments

    J. For.

    (2017)
  • J.P. Caspersen et al.

    Seedling recruitment in a northern temperate forest: The relative importance of supply and establishment limitation

    Can. J. For. Res.

    (2005)
  • D.C. Dey

    Sustaining oak forests in Eastern North America: regeneration and recruitment, the pillars of sustainability

    For. Sci.

    (2014)
  • D.C. Dey

    The ecological basis for oak silviculture in Eastern North America

  • Dey, D.C., Fan, Z., 2009. A review of fire and oak regeneration and overstory recruitment, General Technical Report...
  • R.C. Drever et al.

    Can forest management based on natural disturbances maintain ecological resilience?

    Can. J. For. Res.

    (2006)
  • R.T. Fahey et al.

    Shifting conceptions of complexity in forest management and silviculture

    For. Ecol. Manage.

    (2018)
  • Franklin, J.F., Mitchell, R.J.J.R.J., Palik, B.J.B.J.J., 2007. Natural disturbance and stand development principles for...
  • Franzmeier, D.P., Steinhardt, G.C., Schulze, D.G., 2004. Indiana soil and landscape evaluation manual. West Lafayette,...
  • R.P. Guyette et al.

    Dynamics of an anthropogenic fire regime

    Ecosystems

    (2002)
  • M.A. Homoya et al.

    The natural regions of Indiana

    Proc. Indiana Acad. Sci.

    (1984)
  • Cited by (7)

    • Light and regeneration patterns following silvicultural gap establishment in Quercus dominated stands of the northern Cumberland Plateau, USA

      2022, Forest Ecology and Management
      Citation Excerpt :

      Although we pooled regeneration metrics across all azimuthal transect orientations within an experimental unit to evaluate distance-based treatment response, previous research has concluded that oaks demonstrate a varying response to spatial location and transect orientation. Greenler and Saunders (2019) determined that the frequency of established and competitive oak reproduction in 0.4 ha gaps were not only higher in forest matrices than in gap interiors, but were also higher in matrices on the east, north, and west sides of the gap compared with the southern area, which was attributed to intermediate light levels at these locations. In our study, the mean oak seedling height (50 cm) was shorter than more than half of competitor stems within the gap interior.

    • Prescribed fire and natural canopy gap disturbances: Impacts on upland oak regeneration

      2020, Forest Ecology and Management
      Citation Excerpt :

      Alternatively, we may have failed to measure the gap characteristics that predict regeneration density, such as location within the gap or soil moisture. For example, in central and eastern temperate forests like those studied here, most light within canopy gaps is concentrated at the center and northern edge of the gap, which can influence seedling and sapling density (Canham et al., 1989; Greenler and Saunders, 2019). Several studies also noted that differences in gap soil moisture can impact oak regeneration density (Hutchinson et al., 2012a; Iverson et al., 2008, 2017).

    View all citing articles on Scopus
    View full text