Rapid Ecological Integrity Assessment Metrics to Restore Wildlife Habitat and Biodiversity for Shortleaf Pine–Oak Ecosystems

Open woodlands dominated by shortleaf pine (Pinus echinata Mill.) and oak are historically an important component of the landscape across the southeastern United States. These ecosystems support numerous wildlife species, many of which have declined in recent years as the amount and condition of their habitat have declined. Land managers and private landowners need guidance on how to efficiently and accurately quantify the condition and wildlife habitat value of the pine stands that they manage. Here we provide a set of rapid assessment metrics, based on NatureServe’s ecological integrity assessment (EIA) method, to (a) identify exemplary tracts that provide the best habitat for key wildlife species, and (b) monitor restoration efforts to assess progress toward the improved quality of existing tracts. To ensure an ecologically appropriate scaling of metrics, we distinguished six types of shortleaf pine–oak woodland: A.—Interior Highlands shortleaf pine–oak (including A.1—shortleaf pine–oak forest and woodlands; A.2—shortleaf pine–bluestem woodlands); B—montane longleaf pine–shortleaf pine woodlands; C—southern Appalachian pine–oak woodlands; D—West Gulf coastal plain shortleaf pine–oak woodlands; and E—southeast coastal plain and Piedmont shortleaf pine–oak woodlands. We relied on a narrative conceptual model and peer review-based indicator selection to identify a core set of 15 stand-level metrics (two were optional). Individual assessment points (thresholds) and ratings (Excellent, Good, Fair, and Poor) were developed that were sensitive to the distinct attributes of each of the five shortleaf pine–oak and Appalachian pine–oak types. Values for the metrics can all be collected using rapid field methods, such as using basal area prisms and ocular (visual) estimates of cover. Protocols for the consistent application of these EIA methods are provided. A case study is presented from the Cherokee National Forest in Tennessee. These methods provide improved and rapid EIA metrics for all shortleaf pine–oak ecosystems in the southeastern US to help guide conservation-minded landowners in assessing the biodiversity and priority wildlife values of shortleaf pine–oak and southern Appalachian pine–oak ecosystems.


Introduction
The open savannas and woodlands dominated by southern yellow pines are historically a large component of the landscape across the southeastern United States. These pine savannas and open, usually grassy, woodlands are dominated by longleaf (Pinus palustris Mill.), slash (Pinus elliottii Engel.), shortleaf (Pinus echinata Mill.), loblolly (Pinus taeda L.), pitch (Pinus rigida Mill.), and Table Mountain (Pinus pungens Lamb.) pines (taxonomy follows the PLANTS database [1]). They support high plant and animal diversity, with over 900 plant species considered endemic to this and adjacent ecosystems [2], and they are an integral part of the North American coastal plain biodiversity hot spot [3]. They provide several key ecosystem services, including hunting and other outdoor recreation, watershed protection, and high-quality, sustainable forest products, such as poles and pilings, in addition to high biodiversity. However, despite their importance, these open pine ecosystems have experienced many management challenges. As human populations increased, and land management practices and land-use patterns changed, these once-dominant open pine ecosystems were cleared for agriculture and/or development, resulting in significant declines in both the extent and quality of open pine ecosystems across the southeast [4]. In addition, a lack of wildland fire has adversely affected ecosystem structure and function, as well as the flora and fauna that rely on the ecological conditions created by frequent low-intensity fire [5].
Longleaf-dominated woodlands have been the focus of southern pine stand restoration [6][7][8][9][10][11][12] but other open pine stands in the region that are dominated by shortleaf pine, as well as pitch pine and Table Mountain pine in the southern Appalachians, have also experienced substantial declines. Stands of shortleaf pine and shortleaf pine-oak are estimated to have declined by over 90% and have continued to decline since 1980 [5]. These shortleaf pine woodlands provide a habitat for many of the same plant and animal species that depend on longleaf pine ecosystems, further south. However, their ecological characteristics are not the same as those of longleaf ecosystems. The longleaf pine rapid assessment metrics were extended to include other southern open pine ecosystems [9]; there is a need to account for the distinctive features of shortleaf, pitch, and Table Mountain pine-oak woodlands. Recognition of this need led the Southern Region of the Forest Service to support the Shortleaf Pine Initiative, which published a shortleaf pine restoration plan [13]. That plan provided some guidance on assessing ecological conditions but gave no developed set of metrics to guide restoration.
Developing assessment methods of ecological condition necessitates careful consideration of the complexity of terrestrial ecosystems. It is important to develop, test and validate indicators that provide reliable, cost-effective, easily measured, and ecologically meaningful information on ecological integrity. If well-chosen, indicators can also provide early signs of ecosystem degradation and identify areas at risk of ecosystem collapse [14,15]. These methods are broadly referred to as ecological integrity assessment (EIA) methods [16][17][18][19][20][21][22].
Here we propose a set of rapid EIA metrics that can be applied to any of the shortleaf and southern Appalachian pine ecosystems in the southeast, building on the EIA methods previously developed for longleaf pine woodlands [8,9]. Our goal is, in part, to provide the Forest Service with an assessment method that can be applied to all such pine-oak ecosystems of the National Forests of the Southern Region of the Forest Service, as well as other federal, state, private and Native American lands in the region. Specifically, we address three objectives: 1.
To provide a common framework for delineating the various broadly distinct shortleaf pine-oak (SPO) and Appalachian pine-oak (AP) ecosystems. Product: Classification of SPO and AP ecosystems, including distribution according to ecoregion and National Forests.

2.
To define the desired reference conditions that can guide the management of SPO and AP ecosystems, where the primary objective is the conservation of wildlife and biodiversity maintenance. Product: A set of reference conditions for SPO and AP ecosystems, evaluated using ecological integrity indicators. 3.
To provide a rapid EIA assessment protocol that is usable by land managers to assess the condition of stands. Product: A protocol, usable in the field, that guides the evaluation process, using indicators with ratings of Excellent, Good, Fair, and Poor.

Definition of Southern Open Pine Ecosystems
We define SPO and AP ecosystems broadly, in concert with the definition of Southern yellow pine ecosystems that was used to guide the longleaf work [8]: Southern open pine woodlands include large-scale (or formerly large-scale) ecosystems dominated by an open canopy of southern yellow pine trees and used by a great variety of game and non-game wildlife species and plants. Typical dominant species include not only the longleaf pine (Pinus palustris) and shortleaf pine (Pinus echinata) but also pitch pine (Pinus rigida), Table Mountain pine (Pinus pungens), loblolly pine (Pinus taeda), and slash pine (Pinus elliottii). In the coastal plain, longleaf pines historically occurred in extensive large-scale patches, whereas in the Appalachians, pine woodlands often occur in smaller patches on ridges or dry slopes. Historically, these open pine ecosystems contained a diverse ground cover composed of native warm-season grasses and forbs, often with some low shrubs and only sparse tall shrubs. These open conditions were maintained by natural processes, including fires and grazing. These ecosystems are to be found from the West Gulf coastal plain and the Ozark and Ouachita Mountains to the southern Appalachians, Piedmont, the Atlantic and East Gulf coastal plains, and south into the Florida Peninsula.

Priority Wildlife Species
A list of 64 representative wildlife species was developed for the SPO and AP ecosystems (Appendix A, Table A1). Of these, 18 are rare or focal species for this project (Appendix A, Table A2). Based on the literature and an expert review, populations of these species should respond positively in open pine landscapes with improved ecological integrity [6][7][8][9]13].

Management of Southern Open Pine Forests
There is a variety of forest management approaches available to regenerate oak and pine stands, but prescribed fire is very important for limiting the encroachment of fireintolerant trees, and for maintaining the open grass and low shrub-dominated groundcover that is preferred by wildlife [5,13]. Specific management goals will vary with the intent of the landowners and land managers, but may include timber harvesting, including highvalue and small diameter timber (e.g., biomass), the use of herbicides to reduce competition (particularly to control invasive exotic plants), the reseeding of native groundcover species (such as native warm-season grasses, native legumes and other native plants appropriate for the site), mechanical midstory reduction, managed grazing, and other methods.

Purpose
The purpose of these rapid ecological assessment metrics is to help guide ecological restoration forest management decisions in distinct shortleaf pine-oak ecosystems. We demonstrate the usefulness of ecological classification for developing these metrics. We provide methods to repeatably measure ecological integrity and to enable the assessment of changes in condition using the categories of Excellent, Good, Fair, and Poor. As forest management or ecological restoration decisions are implemented, the metrics presented here can enable the assessment of progress over time and can be applied in conjunction with other common timber cruising or wildlife survey methods.

Ecological Integrity Assessment Methods: Metrics-Based Approach
NatureServe and its network partners from various state Natural Heritage Programs, in collaboration with a variety of agency partners, have developed methods to assess ecosystem condition, using the concept of ecological integrity [19][20][21]23]. Ecological integrity can be defined as "an assessment of the structure, composition, function, and connectivity of an ecosystem as compared to reference ecosystems operating within the bounds of natural or historical disturbance regimes" [19,24]. To have integrity, an ecosystem should be relatively unimpaired across a range of ecological attributes and spatial and temporal scales.
Our approach to developing metrics for SPO and AP ecosystems builds on the collaborative process used for longleaf pine metrics [8,25]. Core to the methodology is coherent and consistent conceptual ecological models for specific ecosystem types that identify the major ecological factors and key attributes for which indicators are most needed. The process of modeling and indicator selection leads to a practical set of metrics, by which field measures of the indicators can be collected and rated categorically as Excellent, Good, Fair, or Poor. Indicators can be developed for various data sources (Level 1-remote-sensing data, Level 2-rapid field data, and Level 3-intensive field data), but here we emphasize Level 2 rapid assessment methods. Our goal was to develop field metrics that were easy to apply, fast, repeatable, ecologically relevant, sensitive to conditions, and complementary (not redundant). We followed a six-step process for assessing ecological integrity, as described by others [18,19,21,26,27], namely:
Develop the conceptual models of ecosystem types 3.
Specify the level of assessment 4.
Select the indicators and metrics 5.
Provide assessment tools

Specification of Ecosystem Types
Ecological classifications help managers better understand natural variability within and among ecosystem types, and thus play an important role in helping to distinguish sites that differ across a gradient of conditions and stressors. Given the diversity of SPO and AP types, it is critical to organize existing knowledge about their location in the landscape, how they function, and how management decisions affect those functions. We can then organize this knowledge into a conceptual model or descriptive summary of how the ecosystem functions (see Step 2,below).
Shortleaf pine is the most widespread pine species in the eastern United States, occurring in 22 states, with a range of over 1,140,000 square km (440,000 square miles) [28]. It is found in pine-dominated stands and as a component of mixed oak forests and open woodlands. It can grow on a range of soil types, aspects, geology, and hydrologic gradients [13,29]. To classify this variation into ecosystem types, we used the classification of southern open pine woodlands, as developed for the United States National Vegetation Classification (USNVC), a federal standard for classification of the vegetation of the United States [30], and part of the International Vegetation Classification [25,31]. We emphasized types at the USNVC group and alliance levels, the same level(s) at which longleaf pine assessments methods were developed [8] and the level at which wide-scale mapping of vegetation is now available [32]. We developed five SPO groups and one AP group that served as our base units for developing rapid assessment metrics (Table 1, Figure 1). We briefly describe the ecological characteristics of these groups, in preparation for developing conceptual models.

A.
Interior Highlands Shortleaf Pine-Oak

Develop Conceptional Models
Historic information on natural processes and the natural range of variation (NRV) for SPO groups show that these ecosystems are adapted to frequent, two-to twenty-year return interval fires, needing burning for natural regeneration [33,34]. As with longleaf pines, fire plays a key role in maintaining forest structure and composition. More frequent fires result in an open woodland structure, and an open overstory of either pure pine or, more often, mixed pine and oak. There is a variable shrub layer and, often, a well-developed herbaceous understory of grasses and wildflowers. Fire limits the encroachment of fireintolerant hardwoods and less fire-tolerant conifers (especially the eastern red cedar). Even in closed-canopy forests, infrequent fires maintain shortleaf pine forest types in mixed pine stands [13,29].
Altered fire regimes, in terms of intensity, frequency, and season of burn, have drastically changed the ecosystem of the shortleaf pine forest. Lack of fire allows the establishment of fire-intolerant hardwood species, such as water oak (Quercus nigra L.), sweetgum (Liquidambar styraciflua L.), tulip poplar (Liriodendron tulipifera L.), and red maple [35][36][37]. Shortleaf pine recruitment is reduced due to a lack of fires [38]. In the southern Appalachians, where prescribed fire has been reintroduced, the recruitment of shortleaf pines is often lacking, due to the absence of a seed source from mature trees [39]. Loss of shortleaf pines and an open canopy structure had a negative impact on many wildlife species [13].
The general fire regime, ecosystem structure, and compositional diversity of SPO and AP ecosystems, as described above, share many basic features with longleaf pine woodlands, and we can largely use the conceptual models of that system, as developed by Loudermilk et al. (2011) [40] and Hanberry et al. (2018) [41], for developing our assessment method ( Figure 2). Using this conceptual model as a guide, we structured our selection of indicators to span the following key ecological attributes. Canopy: • Canopy and midstory trees: indicators are needed that measure both pine and hardwood abundance and distinguish fire-tolerant pines and hardwoods from fireintolerant pines and hardwoods; these measures track the degree to which woodland management and fire regime is maintaining a fire-tolerant tree composition.

•
Canopy cover and size: Indicators are needed to assess the degree to which woodland management and the fire regime maintain an open and mature canopy structure. Shrubs and Herbs: • Indicators are needed to assess the amounts of tall and short shrub cover, graminoid and overall herbaceous cover, including the cover of invasive plants. Soils: • Indicators are needed to assess damage to the soil profile caused by management and other activities (including past land use).
Visually, our conceptual model is captured by historic images of SPO and AP systems that reflect the open woodland conditions of the model (see photos in Figure 3) and by various restoration efforts already put in place to reintroduce those historic conditions Using this conceptual model as a guide, we structured our selection of indicators to span the following key ecological attributes. Canopy: • Canopy and midstory trees: indicators are needed that measure both pine and hardwood abundance and distinguish fire-tolerant pines and hardwoods from fireintolerant pines and hardwoods; these measures track the degree to which woodland management and fire regime is maintaining a fire-tolerant tree composition. • Canopy cover and size: Indicators are needed to assess the degree to which woodland management and the fire regime maintain an open and mature canopy structure.

Shrubs and Herbs:
Forests 2021, 12, 1739 8 of 32 • Indicators are needed to assess the amounts of tall and short shrub cover, graminoid and overall herbaceous cover, including the cover of invasive plants. Soils: • Indicators are needed to assess damage to the soil profile caused by management and other activities (including past land use).
Visually, our conceptual model is captured by historic images of SPO and AP systems that reflect the open woodland conditions of the model (see photos in Figure 3) and by various restoration efforts already put in place to reintroduce those historic conditions ( Figure 3).

Specify the Level of Assessment
To develop an effective, rapid condition assessment metric approach, our protocols needed to be:  To develop an effective, rapid condition assessment metric approach, our protocols needed to be:

•
Able to distinguish condition/ecological integrity of remnant stands and restored stands using metrics that are, where necessary, sensitive to particular types of open pine ecosystems • Repeatable, based on clearly defined protocols • Quick to apply in the field, with multiple locations assessed in a site visit of 2-4 h • Relatively simple and easy for field crews to implement • Able to detect change over time (for monitoring) • Contribute information to regional assessments.
We developed indicators that met these criteria as follows: Canopy-The abundance of southern yellow pine and hardwood can be measured using either (a) estimates of the basal area from 2 × BAF (basal area factor) metric prism or 10 × BAF English prism from at least four locations within an assessment area, or (b) direct measures of all stems that are 12.7 cm (5") diameter or greater at 1.37 m (54"), diameter at breast height (DBH) in centimeters. In both cases, measures are converted to basal area measures in m 2 /ha (ft 2 /acre). Basal area metrics are documented using metric system (SI) units, but the English units are also provided as they are commonly used by managers within the United States.
Fire tolerance-To ensure consistency in the indicator process, we developed a list of tree species that are considered largely fire-tolerant or fire-intolerant (see Nordman et al., 2021) [42]. This was achieved through a literature review, and with expert input. Factors contributing to hardwood fire tolerance include: • Bark resistance to fire • Ability to resprout after fire • How well any downed leaves burn in fuel beds • How fire influences recruitment • Regular occurrences of fire in stands that have been frequently burned over a long period of time.
Composition-Emphasizing tree species identification. Researchers should rely on growth form categories for shrubs and herbs, but the option of collecting species composition data should be provided, given the widespread availability of floristic quality assessment (FQA) methods. The one exception is that invasive plant species need to be distinguished from noninvasive species.
Soils-Reliance on visual (ocular) estimates of soil disturbance.

Select the Indicators and Metrics
Metrics specify both the measures needed to quantify the indicators, and also the rating scale by which those measures indicate the levels of ecological integrity. To guide the selection process, the following criteria were used for each metric: (a) representing a key ecological attribute or ecosystem service (from the conceptual model), (b) not being too noisy/variable (i.e., both short and long-term trends should be detectable), (c) not being prone to measurement error, (d) cost-effective, (e) feasible (including the speed and ease of measurement in the field), (f) relevant to management objectives, and (g) commonly being collected.

Determine Metric Assessment Points/Thresholds
Using knowledge of the natural range of variation (NRV) as a guide to the reference condition, we can approximate both the natural variation in a metric and the variation caused by stressors. In this step, we established ecological "assessment points" that distinguished expected or acceptable conditions from undesired ones that warrant further evaluation or management action (see Bennetts et al., 2007) [43], regarding "assessment points" versus "thresholds" as guides for assessing the ecosystem's condition). We use four rating categories for condition (Excellent, Good, Fair, and Poor) for each of the metrics, using assessment points that can be readily measured as part of a rapid assessment protocol. We conducted a literature review for insights into the factors driving changes in the various indicators of ecological integrity, and characteristic wildlife [42]. We then refined the metric ratings for SPO types that are already partly covered by the Longleaf Pine-Oak (LPO) project [8,9], and we established new metric ratings for the AP. An example using our metric and assessment points approach is shown in Table 2. Table 2. Example of metric and assessment points. The metric and ratings for VMID1: Midstory Fire-Tolerant Hardwood Cover, as developed for the Southern Appalachian pine-oak group, allows for a greater proportion of fire-tolerant oaks under natural conditions than for the southeast group. We developed definitions for the metrics, shown in Table 3, and summary tables (Appendix B) for each of the six SPO and AP groupings, showing the metrics and their ratings, following the format shown in Table 2 and Table 4. Table 3. Ecological integrity metrics to assess the shortleaf pine-oak and Appalachian pine-oak groups. Fifteen stand metrics are provided, two being optional, for the canopy, midstory and shrub, herbaceous cover, and soil.

Metric
Metric Basal area ≥ x m 2 /ha of y pine trees ≥ z" DBH class An assessment of amount of cover of tall shrubs (0.9-3.0 m tall) and small broad-leaved trees less than 3.0 m tall.
Tall shrubs average x% to y% cover 3 VSHR2. Short Shrub (<0.9 m tall) Cover An assessment of amount of cover of short shrubs (<0.9 m tall).

GROUND LAYER METRICS
Ground layer consisting of all herbaceous (non-woody) species.

VGRD1. Overall Native Herbaceous Ground Cover
Percentage cover of all (native) herbaceous species in the ground layer.
x% to y% herbaceous cover 3

VGRD2. Native Graminoid Cover
Percentage cover of all native graminoid cover in the ground layer. Native graminoids include grasses and grass-like plants (grasses, sedges, and rushes, including beaked rushes).
x% to y% cover of all native graminoids 3 VGRD3. Floristic Quality Index, Mean C (optional) The Floristic Quality Index is based on the plant species that are present in an assessment area, and their coefficients of conservatism, or C-values. Each species has a C-value, which is an integer from 0 to 10 assigned by botanical experts. High C-values indicate a native plant's fidelity to natural areas with natural processes. Low C-values indicate weedy plants that commonly occur in ruderal habitats, such as old fields or vacant lots. The integer 0 is typically reserved for exotics.
Mean C x to y 1

VGRD4. Invasive Plant Presence/Distribution
Percentage cover of all invasive species in all layers.
Invasive non-native plant species, x% to y% cover 1 Protocols are also needed to ensure that consistent and clear guidance is provided to field crews on how data are gathered for each metric. We developed protocols that included the following key pieces of information: (1) definition of the metric, (2) rationale for the selection of the metric, (3) measurement protocol, (4) metric ratings, (5) data for metric ratings, (6) scaling rationale, (7) citations.

Peer-Review Workshops to Refine the Metrics
Peer review meetings were held with the Forest Service staff and external partners in January and February 2021. In these meetings, we conducted a systematic review of all metrics, including an evaluation of the assessment points. In the introductory meeting, we presented the need for monitoring of these ecosystems, the background, characteristic wildlife, rapid EIA approach, and goals in the development of the rapid assessment metrics for SPO and AP ecosystems. In each of the three subsequent meetings, we reviewed the metric ratings for the various SPO and AP groups in a specific region: namely, the Interior Highlands (Ozark and Ouachita Highlands, Groups A.1 and A.2), the Montane and Southern Appalachian (Groups B and C) and the Coastal Plain and Piedmont (Groups D and E). All review comments were tracked, and a formal peer review response form summarizes how we addressed these comments (available on request from J.B.).

Case Study of EIA Metrics
To advance our understanding of the use of metrics, we evaluated the condition of stands, utilizing an existing dataset from the Cherokee National Forest in eastern Tennessee. Here, and across the Southern Region of the Forest Service, plot data is collected to measure the effects of prescribed fire [44,45]. We selected 11 sites that correspond to the AP group where shortleaf or pitch pine stands were being managed and monitored. Stands varied according to the levels of oaks and southern yellow pines present. Each site has a different fire history related to frequency and seasonality, with 6 sites experiencing some prescribed burning and 5 sites that did not.
We used the relevant plot measurements (basal area, canopy cover, species, etc.) to assign a score based on metrics variants that were appropriate for the AP group. Absolute cover was converted to canopy cover, based on Jennings et al. (2009) [46] and, for 2 midstory metrics, the basal area was used to estimate canopy cover. The monitoring protocol generally specifies data collection for 10 of the 14 metrics; 4 canopy, 4 midstory/shrub, and 2 ground layer metrics; however, shrubs (2 metrics) and ground cover (2 metrics) were not consistently collected because some portions of the fire-effects monitoring protocol are optional, based on the objectives of the prescribed burn. Thus, only 6 metrics were sufficiently measured across the plots. The presence of invasive species, forest soil disturbance, and the canopy cover of southern yellow pine were not scored for all plots.

Ecosystem Types
Our initial set of SPO and AP types provided a suitable framework for guiding our development of metrics. One refinement to the Interior Highlands shortleaf pine-oak group (A) was to split it into two groups: A.1, Interior Highlands shortleaf pine-oak Forest and woodlands (IH1); and A.2, Interior Highlands shortleaf pine-bluestem woodland (IH2). This refinement was made early in the process and reflects the very different site conditions and fire regimes that shape these two groups. Another refinement was to split West Gulf coastal plain pine-oak woodlands (WG) from the southeast coastal plain and Piedmont shortleaf pine-oak woodlands (CP) because in the West Gulf coastal plain group, loblolly pine and shortleaf pine are both important canopy-dominant trees. These distinctions are also recognized at the USNVC group or alliance levels ( Table 1).

Metrics List
Our peer review-based process identified 15 stand metrics (Table 3). For the stand metrics, there are 6 canopy metrics, 4 midstory/shrub metrics, 4 ground layer metrics, and 1 soil metric. A brief definition and the measures used for each metric are provided in Table 3 and in Appendix B. Full documentation for each metric can be found in the study by Nordman et al. (2021) [42].
As indicated by the conceptual model, the relative roles of pines and hardwoods are important indicators of these ecosystems, and the canopy and midstory metrics emphasize these attributes. Both VCAN4-Canopy Fire-Intolerant Tree Basal Area and VMID2-Midstory Fire-Intolerant Tree Cover include both hardwoods and conifers (the longleaf metrics only included hardwoods) because several conifers, such as red cedar (Juniperus virginiana L.) and eastern white pine (Pinus strobus L.) are relatively fire-intolerant in the region where SPO and AP groups are found. Overall stand structure is assessed via VCAN3-Southern Yellow Pine Stand Size Structure and VCAN6-Stand Density Index. The stand density index is optional because it is a complex metric to interpret, and it has not yet been developed for AP. Thus, five main metrics are developed to characterize the canopy layer.
Shrub and herb layer indicators are used as good short-to medium-term responders to fire regimes, and thus help interpret fire dynamics. VGRD2-Native Graminoid Cover is broader than the equivalent LPO Native Warm Season Grass metric because a mix of cool and warm season grasses occur in many SPO and AP groups. VGRD3-Floristic Quality Index, Mean C replaced the LPO Native Wiry Graminoid Cover metric because it more comprehensively assesses all species. It does require greater botanical expertise and thus is an optional metric, but the metric has been widely applied in many condition assessments, such that we encourage its development and use. SDIS1-Forest Soil Disturbance assesses the soil condition directly, rather than the LPO usage of plant indicators to track soil disturbance (Herbaceous Indicators of Soil Disturbance). Our metric both reduces the botanical expertise required for the metric and is a more direct measure of the key ecological attribute being measured.

Metric Assessment Points
The metric assessment points, varied by SPO and AP types, reflective of their different biological and ecological characteristics, are shown in Table 4. Table 4. Metric assessment points for the various SPO and AP groups. Groups that share the same metric assessment points are assigned the same variant number (e.g., IH, ML and AP groups all have the same values for VCAN1, so they are assigned to variant 1. IH = Interior Highlands shortleaf pine-oak. IH1 = Interior Highlands shortleaf pine-oak forest and woodlands; Interior Highlands shortleaf pine-bluestem woodland; ML = montane longleaf pine-shortleaf pine woodlands; AP = Southern Appalachian pine-oak woodlands; WG = West Gulf coastal plain pine-oak woodlands; EP = southeast coastal plain and Piedmont shortleaf pine-oak woodlands.

Canopy
Generally, reference (excellent) conditions include both a dominant pine component and a smaller amount of fire-tolerant hardwoods, such that pines contribute at least 25% cover, and broadly have a 2:1 ratio of fire-tolerant southern pines to fire-tolerant hardwoods, especially oaks. However, Interior Highlands shortleaf pine forest and woodlands contain a higher proportion of fire-tolerant hardwoods. Poor conditions are indicated whenever fire-intolerant hardwoods are three-to four times more abundant than southern pines.
West Gulf coastal plain (WG) and southeast coastal plain-Piedmont (CP) woodland types showed differences in their dominant pines but otherwise prompted the same assessment points across all metrics, sometimes distinct from the Interior and Appalachian woodlands (Southern Pine abundance in VCAN1 and VCAN2) and sometimes similar. Similarly, Interior Highlands had the same assessment points as montane longleaf pineshortleaf pine woodlands (ML) and AP, sometimes shared with the coastal plain groups. ML had a distinct metric variant (VCAN3.v2) related to the distinctive character of montane longleaf pine ecosystems; namely, the flat-topped crowns in older mature longleaf trees. AP has a metric variant related to the smaller-diameter trees of these pines compared to the others. Thus, the metric ratings were sensitive to the distinct canopy characteristics of each group.

Midstory/Shrub
In general, SPO and AP groups have very open midstories, in reference (excellent) conditions. As with the canopy, the midstory patterns were broadly consistent among the groups, with midstory fire-tolerant hardwoods having no more than 2-10% cover when in excellent condition, and greater than 35%-50% when in poor conditions. Fire-intolerant hardwoods also prompt low percentages when stands are in excellent condition, and thresholds for poor conditions are 20%-30% cover.
Shrub metrics vary much more widely from group to group. Tall shrub cover for excellent condition varies from <5%-10% (most groups) to as much as 30% (Interior Highlands shortleaf pine-oak forest and woodlands, IH1), and short shrub cover varies from <20% (most groups) to less than 50% (AP). Similarly, poor conditions are typically encountered when tall shrub cover exceeds 25%-30%, but this cover could be as high as 50% in IH1. AP stands out as largely having a dominant low shrub heath layer in reference conditions, rather than graminoid, herbaceous, or tall shrub layers, with as much as 50% of low shrub cover being part of (excellent) reference conditions, with poor conditions only occurring when the low shrub heath cover exceeds 80%.

Ground Layer
As with the shrub layers, herbaceous and graminoid cover varies from as little as 15% in AP to as much as 35%-45% or more in the SPO groups. As noted above, AP stands out as largely having a heath rather than a graminoid or herbaceous understory. The various SPO and AP groups shared metric variants in various combinations, reflecting the individual responses of their key ecological attributes. However, as with the Canopy and Midstory metrics, the WG and CP always shared metric variants, and the Interior Highlands shortleaf pine-bluestem woodland IH2 and ML often shared the same ground layer metric variant.

Metric Summaries by Open Pine Group
As part of our collaborative process to create metrics, we determined that various combinations of each SPO and AP group varied enough to justify its own set of metrics. The metrics are summarized for each of the six habitat groups in Appendix B. The full documentation of these metric protocols [42] is available for download from Nature-Serve at: https://www.natureserve.org/projects/developing-rapid-assessment-metricsmeasuring-open-pine-ecosystem-health-southeastern-0 (accessed on 12 November 2021).
Our effort to develop rapid assessment metrics resulted in 15 stand metrics. Together, these 15 metrics serve as the best indicators of ecological condition for SPO and AP woodlands. When taken together, these indicators can help land managers and other interested parties understand the ecological condition of their open shortleaf pine-oak stands. These 15 metrics are in four subsets, representing the canopy, midstory, ground layer, and soils. This approach of grouping metrics by strata allows users to assess the condition of the various layers separately [47].

Case Study
To advance our understanding of the use of metrics, we evaluated the condition of stands, utilizing an existing dataset. We selected 11 sites on the Cherokee National Forest where shortleaf or pitch pine was present. The overstory composition included oaks and southern yellow pines; this method may have selected plots that may not be dominated by southern yellow pines. For the analysis, 5 sites were selected (Table 5) with measurements prior to fire (baseline) or controls (no fire management), measured during 2005-2008. For sites with prescribed fire, the 6 plots were (1) measured between 2004 and 2006 after 4 to 7 years, after one prescribed fire, and (2) measured between 2016 and 2019 and had at least one additional prescribed fire. Each of the 6 metrics was individually scored and used to determine the overall stand-level condition score. Table 5. Case study showing metric ratings for Cherokee National Forest, using the southern Appalachian pine-oak woodlands. Six metrics (see Table 3 for descriptions) assessed sites managed with prescribed fire (n = 6) between two time periods and baseline sites (n = 5). For sites with prescribed fire, the plots were (1) measured after 4 to 7 years, after one prescribed fire between 2004 and 2006, and (2) measured between 2016 and 2019, after at least one additional prescribed fire. Five sites were selected with measurements prior to fire (baseline) or controls (no fire management), measured between 2005 and 2008. Evaluation scale: 4.0 to 3.5 = excellent (dark green), 2.5 to <3.5 = good (light green), 1.5 to <2.5 = fair (yellow), <1.5 = poor (red) [8]. The preliminary analysis shows that the metrics helped characterize existing condition in order to evaluate sites. For the 5 sites evaluated with one prescribed fire history, the overall condition score was the same as sites with no fire (2.9-Good vs. 2.9-Good). For the 6 sites with two prescribed fires, the overall condition score did improve (to 3.2) due to the improved score in both the fire-tolerant hardwood basal area (VCAN4 1.4 to 2.3) and fire-intolerant trees metric (VCAN5 1.6 to 2.0); that is, fire helped remove the fire-intolerant tree species.

Metrics-Based Approach and Reference Conditions
The metrics-based approach, used here to assess ecosystem condition, uses a welltested EIA methodology for screening and vetting the merits of each metric, from reviewing source literature to peer review and field-testing. Our goal was to build on expert knowledge from past work on longleaf pine woodlands [8] and from existing expertise across the SPO and AP regions. The six-step protocol for developing metrics proved effective for guiding peer reviewer input on all metrics. Our team used the information and viewpoints gathered from these interactions to revise the draft metrics and produce the final assessment method.
There is concern that the current ecological conditions of ecosystems are changing so rapidly that reference conditions based on the natural range of variation (NRV), or historical range of variation (HRV), as conducted here, may no longer be relevant to our assessment of current conditions. However, there are various ways in which NRV remains an important guide [48,49]:

1.
First, it is the knowledge of natural variability that informs our goals and evaluations of current conditions, but this knowledge does not a priori constrain how we state desired conditions for excellent or good ecological integrity or the level of ecosystem services.

2.
Second, to suggest that we can simply take over the management of natural ecosystems without understanding NRV is to invite failures in these complex systems.

3.
Third, the purpose of understanding NRV is not to lock us in the past but to ensure that we connect the historical ecological patterns and processes to the present and future.

4.
Fourth, understanding NRV will ensure that we can anticipate change and emphasize resilience in the face of future changes.
Our models and our understanding of the NRV of SPO and AP can also be informed by sites that represent reference (excellent) conditions. As described by Brooks et al. (2016) [50], reference sites represent those areas that are intact or with minimal human disturbance, i.e., "reference standard" or "exemplary ecosystem occurrences." Typically, the initial approach to identifying reference sites is to rely on a combination of factors, including naturalness, apparent ecological integrity, and lack of evidence of human disturbances, leading to degradation. Naturalness and integrity are often judged by historical fidelity (connectivity in time), a full complement of native species, characteristic species dominance and productivity, the presence of typical ecological processes such as fire, flooding, and windstorms, and minimal evidence of anthropogenic stressors [51]. The model, attributes, and metrics that document reference conditions can be adjusted through adaptive management feedback loops.
Although not described here, we also recommend assessing the landscape context, an important part of assessing ecological condition. The four landscape-scale metrics we recommend are absolute patch size, contiguous natural land cover, land use index, and perimeter with natural buffer [19]. These metrics help us distinguish between areas that may have high levels of integrity at a smaller scale but may not sustain priority wildlife long-term because of their isolation, proximity to developed areas, or small stand size.

Wildlife and Ecological Integrity Assessments
Open pine habitats, especially those dominated by longleaf pine or shortleaf pine, provide the last refuge for many at-risk and declining vertebrates and many more at-risk and declining plant species. A few species that depend on the open stand conditions being represented by excellent to good condition include the red-cockaded woodpecker (Picoides borealis), Bachman's sparrow (Aimophilus aestivalis), northern bobwhite (Colinus virginianus), brown-headed nuthatch (Sitta pusilla), pine warbler (Setophaga pinus), prairie warbler (Setophaga discolor), and others (Appendix A). Our approach provides regional managers with methods for more accurately documenting progress toward their goals for increasing the area of excellent to good stands of open pine woodland habitat [42,[52][53][54]].

Application of Metrics
The method provided here is for rapid, stand-level applications and, generally, can be applied at sets of points or small plots across stands, in a manner similar to a timber cruise. To implement these rapid assessment metrics, users must first choose the SPO or AP group (the open pine ecosystem type) that best fits the area they are managing and want to evaluate. This choice could be based on one of two situations: (1) the area of interest is currently one of the SPO or AP groups and the manager wants to know its current condition; or (2) the manager wishes to restore one of these SPO or AP groups in an area that has been degraded and whose current land cover is not open pine. For the latter situation, managers will need to work with available potential site models, such as the BioPhysical setting models of LANDFIRE, or the terrestrial ecological inventory units of the Forest Service.
The data presented herein is not intended to be regulatory or administratively prescriptive, nor to conflict with any manager's ability to meet their underlying legislative mandates. As the data and recommendations put forth here reflect the contemporary collective expertise of many ecologists, foresters, biologists, and researchers, we encourage partnerships to iteratively update and refine these data and recommendations, as we increase our knowledge and understanding of wildlife species habitat needs and management strategies within open pine ecosystems across the southeastern United States. With sufficient investment in the assessment of current stand conditions and the identification of opportunities for restoration, we hope that we can contribute to reversing the decline of shortleaf pine-oak ecosystems across the region.  Acknowledgments: This publication is built on a substantial body of work for longleaf and shortleaf pine-oak ecosystems, completed by NatureServe and partners. We thank the USDA Forest Service for their support of this shortleaf pine assessment method. We also appreciate the peer review comments received from researchers and land managers throughout the Southeast and southern Appalachian regions, as documented by Nordman et al. (2021). In particular, the scientists of the Forest Service, Southern Research Station and scientists at Land-Grant Universities have helped greatly improve our understanding of these open pine-oak ecosystems. We also thank the South Atlantic Landscape Conservation Cooperative, an affiliate of the US Fish and Wildlife Service and other conservation partners, in supporting open pine assessments in the region. NatureServe is a network-based organization, and our strength in the US comes from the Natural Heritage Programs, many of whom have worked with NatureServe to advance these ecological integrity assessment methods. We would also like to thank the reviewers who provided feedback on the earlier versions of this manuscript. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government. USDA is an equal opportunity provider, employer, and lender.

Conflicts of Interest:
The authors declare no conflict of interest. Table A1. Representative wildlife species of shortleaf pine-oak and Appalachian pine-oak ecosystems Focal and rare taxa are in bold. For each taxon, the NatureServe Global Rank (GRank) is provided (G1 = Critically Imperiled, G2 = Imperiled, G3 = Vulnerable, G4 = Apparently Secure, and G5 = Secure). Subspecies ranks are shown as T ranks (e.g., G5T1). The Endangered Species Act (ESA) listing includes EN = Endangered, TH = Threatened.   Table A2. Focal and rare species of shortleaf pine-oak and Appalachian pine-oak woodlands. SGCN = species of greatest conservation need.

Appendix B. Metric and Metric Ratings for Each Shortleaf Pine-Oak Group and Southern Appalachian Pine-Oak Group
Here we summarize the metric ratings for each shortleaf pine-oak or Appalachian pine-oak group. Please refer to Nordman et al., 2021 for more detailed information on each of the metrics.