Ecological Archives A025-015-A2

Gabriel I. Yospin, Scott D. Bridgham, Ronald P. Neilson, John P. Bolte, Dominique M. Bachelet, Peter J. Gould, Constance A. Harrington, Jane A. Kertis, Cody Evers, and Bart R. Johnson. 2015. A new model to simulate climate-change impacts on forest succession for local land management. Ecological Applications 25:226–242. http://dx.doi.org/10.1890/13-0906.1

Appendix B. STM cover type descriptions.

Ponderosa pine (Pinus ponderosa Lawson & C. Lawson) and Pacific madrone (Arbutus menziesii Pursh)occur in current forests in our study area in scattered locations at low abundances but are never community dominants. We describe each current and future community type using species that already occur in the study area and appear to be likely potential components of future communities. The distribution of Pacific madrone includes subtropical climate zones, and Ponderosa pine is representative of a conifer in a continental climate. A unique feature of our study area is its location near the confluence of three different climatic regions: maritime temperate, continental temperate, and subtropical. This location makes the increasing prominence of madrone and ponderosa pine plausible over the coming century.

In order to rectify the often-conflicting spatial distributions and descriptions of vegetation in the available data sets, we used a logical rule-set that gave us one state for each IDU on the landscape. We created a 30-m grid representation of each data source and then prioritized them to assign a vegetation state to each pixel based on our on-the-ground knowledge of which were most accurate in particular situations [1–4] and consultation with regional ecologists. We then aggregated the grid cells into their respective IDUs, and assigned to each IDU the plurality state of its constituent 30-m cells. We validated our results for several subareas of the study area using detailed field data from The Nature Conservancy that had been mapped in GIS.

To assign states to the initial landscape, we collected descriptions of vegetative cover from multiple data sources, including the Gradient Nearest Neighbor [5], Oregon Gap Analysis Program [6], Land Use Land Cover [7], Northwest Habitat Institute [8], 1851 Vegetation [7]. These sources were necessary to generate lists of trees, grouped by species, size class, canopy closure, and canopy layering, that describe extant vegetation, especially prairie and oak savanna habitats of limited current distribution. Although the Gradient Nearest Neighbor data set was the only one that specified each of the required four parameters, it was not developed for the fine-scale vegetation assignments that we implemented in this project and required substantial reclassification from other data sources to better reflect existing vegetation.

QMD classes were: no QMD, trees < 1.37 m tall; 0-12.7 cm diameter at breast height (DBH, 1.37 m); 12.7-25.4 cm DBH; 25.4-50.8 cm DBH, and >50.8 cm DBH. Canopy closure categories were < 25%, 25-60%, and > 60%. We defined the canopy to have either one or two layers. (STM development is described in greater detail in the supplemental materials.)

OA: Open broadleaf deciduous communities of drought-tolerant species, typically oaks. May include other related genera, but must have canopy cover below 25%. This group includes most prairie and savanna.

OW: Broadleaf deciduous woodland of principally drought-tolerant species, typically oaks. May include other related genera. Must have canopy cover between 25% and 60%.

OD: Woodlands of drought-tolerant trees, dominated by broadleaf deciduous trees rather than needleleaf evergreen trees. Trees must have quadratic mean diameter (QMD) of at least 25.4 cm. If the QMD is less than 25.4 cm, these communities are usually described as DO.

DO: Woodlands and low-density forests of needleleaf evergreen trees growing above drought-tolerant broadleaf deciduous trees.

DD: Less mesic needleleaf evergreen woodlands and forests. These may contain a wide variety of species, but Douglas-fir typically dominates.

BM: Mesic broadleaf deciduous forest, bigleaf maple usually dominates in upland locations but may include species of alder, cottonwood and ash in riparian zones. This may include a substantial component of mesic needleleaf evergreen trees.

DM: Mesic mixed needleleaf evergreen and broadleaf deciduous forest. The typical needleleaf evergreen species is Douglas-fir, but there may be a grand fir component. The typical broadleaf deciduous species is bigleaf maple. The needleleaf evergreen component must be dominant over the broadleaf deciduous component.

DG: Mesic needleleaf evergreen forest. Douglas-fir is the dominant species, with grand fir as the subdominant species. There may also be substantial quantities of bigleaf maple.

M: Systems dominated by evergreen broadleaf species, typified by madrone. This must not include a substantial Douglas-fir component.

MD: Systems dominated by evergreen broadleaf species, with a substantial component of Douglas-fir .

P: Systems dominated by xeric evergreen species, typified by ponderosa pine. This includes prairie, savanna and woodland systems.

The ctss (cover type, structural stage) description of an STM state is a concatenation of five components.

First is the cover type:

Next is size class:

Canopy closure is next, although it is only included for size classes p, s and l:

An exception to this rule is the dd cover type, for which open is defined as ≤ 60% canopy cover, and medium canopy closure is > 60% canopy cover.

Next is the canopy layering, included only for sizes s and m:

Finally, there may be "rf" appended, indicating a managed state with reduced fuels, or a "p" indicating a post-disturbance state that persists for only a single year to allow PVT effects on stand regeneration to be implemented. In its current implementation, the STM in CV-STM includes a total of 111 unique states, although the current design could support up to 396 states.

Literature Cited

1. Yospin GI, Bridgham SD, Kertis J, Johnson BR (2012) Ecological correlates of fuel dynamics and potential fire behavior in former upland prairie and oak savanna. For Ecol Manag 266: 54–65.

2. Day JW (2005) Historical Savanna Structure and Succession at Jim's Creek, Willamette National Forest, Oregon Eugene, OR, USA: University of Oregon.

3. Sonnenblick KL (2006) Environmental Controls Over Forest Succession of a Former Oak Savanna, Jim's Creek, Willamette National Forest, Oregon [Masters]. Eugene, OR, USA: University of Oregon.

4. Murphy MS (2008) Edaphic Controls over Succession in Former Oak Savanna, Willamette Valley, Oregon [Masters]. Eugene, OR, USA: University of Oregon.

5. Ohmann JL, Gregory MJ (2002) Predictive mapping of forest composition and structure with direct gradient analysis and nearest- neighbor imputation in coastal Oregon, U.S.A. Can J Res 32: 725–741. doi:10.1139/x02-011.

6. Existing Vegetation: NW ReGAP (n.d.). Available: http://pdx.edu/pnwlamp/existing-vegetation. Accessed 15 November 2011.

7. Hulse DW, Gregory SV, Baker J (2002) Willamette River Basin Planning Atlas: Trajectories of environmental and ecological change. Corvallis, OR, USA: Oregon State University Press.

8. Northwest Habitat Institute (n.d.). Available: http://www.nwhi.org/index/gisdata. Accessed 15 November 2011.

[Back to A025-015]