Vegetation Quality Assessment: a sampling-based, loss-gain accounting framework for native, disturbed and reclaimed vegetation

Vegetation Quality Assessment: a sampling-based, loss-gain accounting framework for native, disturbed and reclaimed vegetation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Vegetation Quality Assessment: a sampling-based, loss-gain accounting framework for native, disturbed and reclaimed vegetation Bradley L. Boyle, Warn Franklin, Alison Burton, Raymond E. Gullison This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3374961/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Jan, 2024 Read the published version in Ecological Indicators → Version 1 posted You are reading this latest preprint version Abstract Governments and society increasingly are demanding that industrial projects result in a net positive impact (NPI) on biodiversity. Impacts are commonly measured in terms of losses and gains of area and quality of vegetation, where quality refers to how closely a site matches the condition of native vegetation in its undisturbed state. Existing vegetation quality frameworks share a number of limitations, including little or no replication, uncertain scope of inference, vulnerability to bias, and inability to measure error. Here we present the Vegetation Quality Assessment (VQA) framework, a sampling-based extension of Quality Hectares that measures vegetation quality in terms of overlap between the probability distributions of ecological indicators at a project site and in undisturbed (benchmark) vegetation of the same kind. Distribution overlap incorporates natural variation at the landscape scale and provides an intuitive measure of quality that varies between 0 and 1. Indicators are measured using a stratified-random sampling design that minimizes bias and supports inference at the scale of the project landscape. Confidence limits of quality and quality hectares are determined by bootstrapping; power and minimum sample sizes are estimated by Monte Carlo simulation. Multiple assessments track losses and gains of quality hectares and enable accurate accounting of progress to NPI. The VQA framework can be implemented using a variety of vegetation sampling methods, allowing existing vegetation databases to be leveraged as sources of data. We conclude by demonstrating the application of VQA at several mining operations in the Elk Valley of southeastern British Columbia, Canada. Environmental assessment ecological monitoring vegetation restoration reclamation loss-gain accounting biodiversity offset Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Increasing global awareness of the pervasive effects of human activity on the natural world (Ceballos et al. 2015 ; Pimm and Joppa 2015 ) is leading to voluntary and mandated approaches to reduce the impacts on biodiversity from resource development and related economic activities (International Finance Corporation 2012 ; Rainey et al. 2015 ; CBD 2020 ). The extent and condition or quality of native vegetation is the principal currency by which impacts to terrestrial ecosystems are measured (Gibbons and Freudenberger 2006 ). In addition to their inherent value as major components of the world's biodiversity (Kier et al. 2005 ), plants and plant communities provide the habitats and ecosystems upon which other organisms depend, and deliver essential environmental services through their influence on soil fertility, hydrological cycles and the global carbon balance (Schlesinger and Andrews 2000 ). In the context of environmental impact assessment, inaccurate estimates of vegetation extent and quality can lead to reduced restoration effectiveness at impacted sites, mismatched biodiversity offsets, and potentially costly compliance failures (Maron et al. 2012 ). An accurate, objective and cost-effective framework for measuring vegetation quality is arguably one of the most important tools of regulators, developers, reclamation practitioners, and other stakeholders involved in monitoring and managing the effects of industrial development on terrestrial biodiversity (Parkes and Newell 2003 , Gibbons and Freudenberger 2006 ). In many assessment and monitoring frameworks, vegetation "quality" refers to the degree to which vegetation at a site resembles native vegetation in the absence of human disturbance (Gibbons and Freudenberger 2006 ). Vegetation structured by natural disturbances (such as flooding and avalanches) or maintained by pre-European human intervention (such as fire-managed grasslands in North American and Australia) are also generally regarded as native vegetation (Parkes and Newell 2003 ; Gibbons and Freudenberger 2006 ; Lewis et al. 2018 ). Although the value of native vegetation per se is uncontroversial (Gibbons and Freudenberger 2006 , Landres et al. 2014 ), debate still surrounds the questions of how to measure the condition of vegetation which has been disturbed by human activity (McElhinny et al. 2005 ; Gibbons and Freudenberger 2006 ; Cook et al. 2010 ) and how to select the appropriate reference vegetation to which the disturbed vegetation is compared and restored (Demeo et al. 2018 ). As part of a broader program of biodiversity monitoring and management at Teck Coal Limited’s (Teck) Elk Valley steelmaking coal mining operations in southeastern British Columbia, Canada, we developed an accounting framework for quantifying impacts and improvements to native vegetation in terms of quality (as defined above) and quality hectares (quality x area; Rio Tinto 2008 ; see Sahley et al. 2017 for an example application). The framework focuses on natural ecosystems, with quality measured empirically relative to undisturbed native vegetation, and reflecting not only average conditions but also variation at the landscape scale. In terms of sensitivity, the method needed to be capable of distinguishing major vegetation types and successional (seral) stages, as well as different types of disturbance—all at temporal and spatial scales relevant to large mining operations. Field methods needed to be efficient, cost-effective, and—ideally—compatible with standard protocols to allow the re-use of existing vegetation data. Most importantly, the framework needed to be repeatable, verifiable, capable of estimating uncertainty, and minimally vulnerable to observer bias. Although initially developed in the specific context of Teck's Elk Valley mining operations, the framework and associated methodology is intended to be widely applicable, and is currently in use at a variety of projects involving multiple companies in Canada, USA, Suriname, Chile and Greece, and in vegetation ranging from boreal forest to desert scrub to lowland tropical rainforest. Here we describe this monitoring framework, which we call Vegetation Quality Assessment (VQA). , Existing loss-gain frameworks We identified several existing approaches for quantifying losses and gains in vegetation or habitat quality. These included Habitat Hectares (Parkes and Newell 2003 ), BioCondition (Eyre et al. 2011 ), Biometric (Gibbons et al. 2008 ) and Quality Hectares (Rio Tinto 2008 ), among others. These approaches typically measure vegetation quality by scoring ecological indicators relative to “benchmarks” representing the expected conditions in undisturbed native vegetation. Benchmark values are determined by sampling undisturbed vegetation adjacent to the project site or by reference to standard values for each vegetation or habitat type, as provided by regulatory agencies (e.g., The State of Queensland 2014 ). Multiple indicator scores are combined in various ways to produce an index of overall vegetation quality (Gibbons and Freudenberger 2006 ). Quality is then multiplied by area to provide a discounted measure of the amount and condition of vegetation to be compensated through restoration at the project site and offsets elsewhere. We follow Rio Tinto (Rio Tinto 2008 ) in referring to quality-discounted area as “quality hectares” (QH). All existing methods failed to meet one or more of our key requirements. The most pervasive shortcoming was lack of a statistical framework. Sample plots were often located deliberately rather than at random, and replication, when mentioned at all, was on the order of 1–3 plots per vegetation type, with no explicit justification for the sample sizes used (e.g., Eyre et al. 2011 ). In the absence of replication and randomization at the appropriate temporal and spatial scales, it is impossible to assess the degree to which the samples represent the site as a whole (Hurlbert 1984 ). Furthermore, inadequate replication and lack of randomization increase vulnerability to observer bias; for example, field crews could accidentally or deliberately inflate quality scores by situating plots in unusually pristine sites. Another consequence of lack of replication is the inability to measure error or assess statistical significance. Quality is frequently reported without measures of variance (Parkes and Newell 2003 ; Eyre et al. 2011 ) as are benchmark values provided by regulatory agencies (e.g., The State of Queensland 2014 ). A second shortcoming was the use of custom field protocols specific to each framework, thereby precluding the use of existing vegetation data. In many countries publicly-accessible repositories of vegetation data can be extensive, representing essentially all types of native vegetation (e.g., Peet et al. 2012 ). In the case of Teck's Elk Valley mining operations, the ability to leverage as reference data an existing provincial database of thousands of standardized vegetation plots (British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) 2016 ) was a key goal of our assessment framework. Without this critical resource we would have faced years of expensive data collection—a commitment of time and resources that many companies would find prohibitive. In addition, the sampling methodology commonly employed for vegetation plots within the provincial vegetation database (BC Ministry of the Environment 2010 , 2021 ) was also used by Teck over many years for mapping vegetation within the Elk Valley. The ability to re-purpose such pre-existing data for vegetation assessment would represent a tremendous saving of time and expense. Many of the methods we reviewed omitted important indicators of vegetation condition. For example, BioCondition omit species identifications in favor of “key attributes or surrogates of biodiversity values that can be rapidly measured in the field” (Eyre et al. 2011 ), with the goal of enabling surveys to be "undertaken rapidly by a range of natural resource managers…not just botanical ecologists" (Parkes and Newell 2003 ). However, any potential gains in speed and ease enabled by such simplifications must be weighed against the information lost by ignoring taxonomic composition. Species composition is a fundamental attribute by which vegetation units are defined under most classification systems (e.g., Jennings et al. 2009 ), and indicators based on taxonomic identification are sensitive barometers of forest health, succession and disturbance. For example, small-scale disturbances can leave imprints on species richness that remain detectable for 100 years or more (Curtin 1995 ). Species composition discriminates at fine temporal scales among regenerating forests of different ages (Kappelle 1995 ). Major classes of temperate and tropical vegetation are readily distinguished by taxonomic composition, even at coarse taxonomic levels such as genus and family (Gentry 1988 , 1993 , 1995 ). Furthermore, the gains in efficiency enabled by discarding taxonomic identifications were not readily apparent. For example, the FS1333 Site Visit protocol used extensively in British Columbia (BC Ministry of the Environment 2010 ) and employed for the vegetation plots included in the current study (see example application, below), includes identification and estimation of relative abundance of all plant species in the plot, in addition to structural and soil measurements. Yet the 1–2 hours taken by experienced practitioners to complete a “Site Visit” vegetation survey (BC Ministry of the Environment 2010 ) is comparable to the 2 hours typically required for a BioCondition plot (Eddy et al. 2011 ), which omits most species-level identifications. In developing a vegetation quality framework within the specific context of the Elk Valley, we sought a method that would allow us to use data collected previously as part of Teck’s ongoing vegetation mapping and data collection program, and to leverage as benchmark data the vegetation inventories maintained by the British Columbia Ministry of Environment and Climate Change Strategy (BC ENV) and used to develop the provincial Biogeoclimatic Ecosystem Classification (British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) 2016 ). In addition, as our ultimate goal is to deploy the framework worldwide, we needed a method that would be applicable to all vegetation types, temperate and tropical, and compatible with a wide variety of commonly-used vegetation inventory methods. Overview of VQA The framework we developed combines useful features of existing methods—including quality measured as similarity to undisturbed native vegetation and an accounting workflow based on QH—with a stratified-random sampling design that enables estimation of error and inference at the scope of the entire project. Losses due to project impacts and gains from restoration and offsets are measured relative to a baseline of vegetation quality and area prior to project initiation . Multiple later assessments at project and offset sites track net quality hectares over time, with the ultimate goal of detecting the point at which baseline QH have been exceeded and NPI has been achieved (Fig. 1 ). A key feature of the method is the measurement of quality in terms of overlap between indicator distributions of project site vegetation and in benchmark-quality vegetation of the same type. The approach can be applied to both forested or non-forested vegetation using widely-used vegetation sampling methods. Below, we outline the steps of the VQA workflow, followed by a trial implementation at Teck’s Elk Valley operations. Loss-gain accounting The complete VQA workflow (Fig. 1 ) consists of a series of loss-gain assessments of quality hectares of each vegetation type present at the project or offset site. Quality hectares is calculated as $$QH=Q \text{x} A$$ 1 where QH is quality hectares, Q is vegetation quality and A is the area of that vegetation in hectares. As Q varies from 0 to 1, QH ranges from 0 to A hectares. The workflow begins with a baseline assessment of QH for each class of vegetation prior to project start. At each subsequent assessment, the net change in QH relative to baseline ( QH net ) is determined by subtracting baseline quality hectares ( QH 0 ) from current quality hectares ( QH ) for the same vegetation: QH net = QH – QH 0 (2) Typically, with resource developments such as mines, not all losses at the project site can be restored. These residual impacts will need to be compensated by additional QH gained through conservation or restoration actions at one or more offset sites (Fig. 1 a). Each offset site will require its own set of assessments, including a baseline assessment and separate calculations of net QH. Unlike the project site, offset assessments generally need to be adjusted for one or more counterfactual scenarios, such that only additional QH directly attributable to the offset program are included. In other words, existing QH that would have persisted into the future, or increases in QH that would have occurred in the absence of the offset program, must be excluded. The simplest counterfactual is complete elimination of an existing natural area. For example, if purchase and protection of an offset prevents a credible scenario of permanent conversion to agriculture resulting in complete loss of all native vegetation, then the baseline quality for the offset would be zero and net offset QH at each assessment would be simply the observed QH. For projects that include offsets, overall net QH is the sum of net QH at the project site and at each offset site. For a project with two offsets, net quality hectares is QH net = QH net.p + QH net.o1 + QH net.o2 , (3) where QH net is overall net QH, QH net.p is net QH of the project site, and QH net.o1 and QH net.o2 are net QH of offset sites 1 and 2, respectively. Each assessment of overall net QH thus includes losses due to project impacts, gains due to restoration and natural succession at the project site, and gains from averted losses and restoration at offset sites. In theory, attaining a NPI requires that QH net > 0 for all native vegetation types by the final assessment. However, each of the many measurements involved in measuring vegetation quality (see "Confidence limits and significance testing") has its own associated measurement error that propagates to all derived values. In VQA, each estimate of Quality, QH and QH net is a expressed as a mean with an associated 95% CI (see "Confidence limits and significance testing"). In practice, therefore, NPI cannot not be assumed until the lower 95% confidence limit of QH net exceeds 0 (Fig. 1 ). Stratification of vegetation, disturbance and reclamation Each assessment cycle requires a comprehensive classification and mapping of all land cover at the project site. Vegetation may be further subdivided into naturally occurring seral stages, if successional vegetation is a major component of natural landscapes in the region. For example, early successional forest can account for large areas in regions subject to frequent, naturally occurring fire (Fulé et al. 1997 ; Korb et al. 2003 ). Each class of vegetation (plus seral stage, if applicable) may be further subdivided into disturbance classes expected to follow divergent successional trajectories. Different restoration treatments should also be monitored separately if recovery outcomes and timelines are expected to differ, or if a goal of the program is to compare differences among treatments. Each class of vegetation plus disturbance (and reclamation treatment, if applicable) represents a different sampling stratum. Within each stratum, the required number of sampling points (vegetation plots; see "Effect size, power, and minimum sample size") should be located at random using GIS software. If the initial classification results in a large number of strata, adequate sample sizes may be difficult or impossible to achieve given available time and resources. In such cases, it may be necessary to reduce the number of strata (thereby increasing per-stratum sample sizes) by combining ecologically-similar vegetation types. Each assessment thus requires its own map of the project site, classified into strata of vegetation, disturbance and reclamation. Once the mapping process is complete, the actual area of each stratum is determined and multiplied by the quality of that stratum (see "Quality calculations", below) to obtain the stratum-specific current QH for that assessment. The QH of all strata sharing the same benchmark vegetation are summed to provide the total current QH for that class of native vegetation. In general, VQA seeks to reestablish vegetation of the same type that existed prior to project impacts; QH, net QH and NPI are therefore assessed independently for each class of benchmark vegetation ("like-for-like"; for a discussion of exceptions, see Text S1). Ideally, the baseline map and baseline assessment are completed prior to initiation of project-related impacts. However, if the VQA framework is implemented after the project has begun, modeling (“back-casting”) can be used to reconstruct vegetation present prior to project impacts. This approach is described in detail in Knopff and Franklin ( 2018 a) and demonstrated in the Teck VQA application outlined below. Vegetation sampling Once mapping is complete and sampling strata have been defined, sampling localities for plots within strata are assigned randomly using GIS software. The number of plots per stratum can be estimated using Monte Carlo methods, as described below. We refer to vegetation plots from the project site as “focal plots”. In addition to focal plots, a reference sample of plots in high-quality, undisturbed examples of each type of native vegetation present at the project site is also required. These reference plots, or “benchmark plots”, can be established in areas adjacent to the project site (if undisturbed native vegetation is present), or obtained from existing vegetation inventory databases. Regardless of source, it is important that benchmark plots meet the following requirements: (a) indicators and sampling methods are identical or compatible with those of the focal plots, (b) the vegetation sampled can be reliably matched to native vegetation types present at the project site, and (c) sampling locations are minimally affected by human disturbance, based on accompanying metadata or expert evaluation. Ideally, experts not involved with the project should be responsible for verifying that plots or localities sampled represent undisturbed native vegetation. In calculating quality, different samples of focal plots representing distinct classes of disturbance or reclamation but the same native vegetation type can be compared to the same set of benchmark plots. The VQA framework is compatible with a variety of vegetation sampling methods. In many cases it may be possible to combine different field methods, as long as indicators measured are compatible. Legacy data collected using different protocols may be used if the relevant indicators can be transformed or otherwise standardized to produce equivalent measurements. A final consideration influencing the choice of plot methods is the indicators to be measured. In most cases, a small number of vegetation attributes—in particular, species richness, taxonomic composition, growth form, and size or age class composition—are sufficient to distinguish among classes of succession and disturbance (Gentry 1988 ; Kappelle 1995 ; Chazdon et al. 2007 ). Most commonly used vegetation sampling methods include such measurements and should be suitable for VQA. Quality calculations Overlap-based quality The principal measure of quality within the VQA framework is overlap between the probability distributions of indicators in the focal and benchmark vegetation. Measuring similarity in terms of overlap between probability distributions has a long history in economics (e.g., comparison of incomes; Gastwirth, 1975 ). The overlap method accommodates a wide variety of normal and non-normal distributions, and deserves to be more widely used—especially in ecology, where skewed and bounded distributions are the norm rather than the exception (Magurran 2004 ). Overlap between focal and benchmark indicator distributions provides an intuitively meaningful index of similarity ranging from 0 to 1, where 1 (100% overlap) means, effectively, "indistinguishable from benchmark". Furthermore, distribution overlap captures the full range of natural variation, including asymmetry not evident in comparisons of means (Fig. 2 ). To calculate overlap, each indicator must first be assigned a distribution family based on the domain of the variable and visual inspection. Appropriateness of the distribution can be formally assessed using a goodness-of-fit test such as the one-sample Kolmogorov-Smirnov test (Berger and Zhou 2014 ). Commonly measured indicators generally fall into one of three distributions: (1) negative binomial (counts such as species richness), (2) gamma (zero-bounded continuous measurements such as taxonomic similarity/dissimilarity), and (3) beta (percentages or proportions such as percent cover of species or growth forms). As the negative-binomial is a discrete distribution, it is approximated by the gamma distribution when calculating distribution overlap (Best and Gipps 1974 ). Overlap is calculated in a two-step process. In step one, separate univariate pdfs are fit to the focal and benchmark samples using maximum likelihood estimation (Fig. 2 a; see Venables and Ripley 2002 ). In step two, the area of the intersection of the two distributions is determined by integration. Confidence limits (CLs) of the overlap-quality estimate are determined by bootstrapping these calculations, using the bootstrap deviance approach described below (see "Confidence limits and hypothesis testing"). Alternative quality methods Although the "overlap quality" method described above accommodates the majority of indicators, three alternative approaches are available to handle special cases. The first of these, one-tailed quality, calculates quality using values either above or below the benchmark mean, but not both. Depending on the goals of the application, one-tailed quality may be appropriate for indicators such as Species Richness if high values of richness are considered desirable; in this case, quality would be calculated using values below the benchmark mean (i.e., the "lower tail"). Details of the one-tailed quality algorithm are provided in Text S2 and Fig. S1 . Fixed-benchmark quality, which applies to beta-distributed indicators only, uses a single fixed benchmark value. Quality is calculated as the absolute value of the difference between the benchmark value and the focal mean, divided by the maximum possible difference on the [0,1] domain, with an additional, non-linear transformation that causes quality to drop steeply with increasing distance from benchmark (Fig. S2). Fixed-benchmark quality is appropriate when a single, ideal benchmark value is readily apparent and appropriate. Examples of indicators for which a fixed benchmark might be used include percent cover of invasive species and infection rates of diseases (e.g., "Percent of trees infected White Pine Blister Rust"); in both cases, the targeted benchmark is 0. Details of the calculation of fixed-benchmark quality are provided in Text S3 and Fig. S2. The third alternative, overlap-means quality, combines features of the overlap and fixed-benchmark algorithms to resolve problems caused by zero and one inflation in beta distributed indicators. Zero- and one-inflated distributions are common among ecological indicators measured as percentages (Magurran 2004 ). Such mixed distributions can distort measures of quality based purely on distribution fitting. The overlap-means algorithm calculates quality twice, once based on distribution overlap and a second time based on the difference between the focal and benchmark mean, and uses a weighted average of the two. The result is a robust quality score which provides an intuitively accurate representation of the similarity of two distributions across the entire beta domain, even for mixed distributions. For details of the calculation of overlap-means quality, see Text S4 and Figs. S3-S4. Overall quality Most existing vegetation quality methods combine indicator scores into a single index of overall quality. To reduce the influence of correlated indicators and minimize subjective weightings (see critique in McCarthy and Parris 2004 ), we group related indicators into “functional groups” representing key aspects of ecological function. For example, indicators based on taxonomic composition, such as species richness or taxonomic composition, are assigned to functional group “Composition”, whereas indicators that measure the relative abundance of different growth forms, such as “percent Cover Herbs” and “percent Cover Trees”, are assigned to functional group “Structure”. For each functional group, functional group quality is calculated as the arithmetic mean of all indicator qualities within that functional group. Overall quality is then calculated as the geometric mean of the functional group qualities, as follows: $$Q= {\left(\prod _{i=1}^{N}Q{f}_{i}\right)}^{\frac{1}{N}},$$ 8 where Q is overall quality, Qf i is the arithmetic mean of indicator qualities in functional group i , and N is the number of functional groups. The use of the geometric mean addresses the criticism that existing approaches are vulnerable to manipulation of quality scores through the substitution of one indicator for another (McCarthy and Parris 2004 ). For example, in quality frameworks that score indicator “downed dead wood” as a habitat feature, it is possible to inflate quality while harvesting all trees if some of trees are left on the ground as "downed dead wood" (McCarthy and Parris 2004 ). Under the VQA approach, a single functional group whose mean drops to zero will force overall quality to zero as well, regardless of the values of the other functional group means. This behavior ensures that all key components of ecological function have non-zero quality. At the same time, use of the arithmetic mean for functional group quality prevents indicators within the same group from forcing overall quality to zero during early succession, when cover (and quality) of herbs and shrubs recovers long before cover (and quality) of large trees becomes non-zero. Confidence limits and significance testing In VQA, the 95% CLs of indicator quality, functional group quality, overall quality and QH are estimated using the bootstrap deviance method of Efron ( 1979 ). Each quality calculation is repeated a large number of times (generally, 10,000 iterations), sampling the actual data with replacement. The lower 95% CL is calculated by subtracting the 0.025 quantile of the bootstrap deviances (i.e., deviance of the bootstrap estimates from the bootstrap mean) from the observed overlap. Likewise, the upper 95% CL is calculated by adding the 0.975 quantile of the bootstrap deviances to the observed overlap. This method accurately represents asymmetric confidence intervals (CIs) that commonly result from overlap of distributions such as the beta distribution. Under certain circumstances, 95% CLs can be used directly to evaluate significant differences. For example, the difference between a measured value of quality and a fixed target value is significant at p ≤ 0.05 if the fixed value lies outside the 95% CLs of the observed distribution. The difference between two estimates of quality or quality hectares is significant if the CIs of the difference of the estimates do not overlap 0 (Cousineau 2017 ). In VQA, we assess progress to NPI by examining the position of the lower 95% CL of the difference between current and baseline QH (i.e., Net QH). Once the lower CL exceeds 0, we can be 95% confident that QH current > QH baseline and NPI has been attained, given the variability of our estimates of Quality and QH (see Fig. 1 b). Effect size, power, and minimum sample size The difference between the mean and the 95% CL, or margin of error (MOE), is an intuitive measure of effect size, precision and error which deserves to be more widely used (Cousineau 2017 ). Because MOEs calculated using the bootstrap deviance method are frequently asymmetrical, we use the largest of the two MOEs of quality as a conservative measure of the "minimum detectable effect size" (MDES) attainable at a given sample size. Defining effect size in this way, we can use Monte Carlo methods to estimate the minimum sample size ( n.min ) required to detect a given target MDES ( MDES exp ) at α = 0.05 and a conventional power of 0.80 (see Text S5 for details). This approach is illustrated in the example application below. Throughout, we follow the recommendations of Cumming ( 2014 ) and Cousineau ( 2017 ) in presenting upper and lower CLs in square brackets. Example application: A VQA accounting of baseline & reclaimed vegetation at Teck Elk Valley Mine Operations Goals of the application As a trial application of the VQA framework, we performed assessments of baseline and current reclaimed vegetation quality for Teck’s steelmaking coal mining operations in the Elk Valley of British Columbia, Canada. The baseline assessment estimated quality and extent of all vegetation at the start of five Elk Valley operations, prior to mining-related impacts. The current assessment measured present-day extent and condition of reclaimed vegetation within the study area. The main goals of the application were four. The first goal was to demonstrate the retroactive application of the VQA framework, using modeling approaches to reconstruct terrain and vegetation originally present within an area already impacted by mining activities (see Knopff and Franklin 2018 for details). The second goal was to develop a best estimate of the baseline extent and condition of all native vegetation impacted by Teck’s operations in the Elk Valley over the period 1950–2018. The third goal was to demonstrate the use of existing vegetation data within the VQA framework by combining focal data from vegetation surveys previously collected from the project site with benchmark data from a public vegetation inventory database. The fourth goal was to estimate the current contribution of reclamation to NPI, with the aspirational goal of ensuring that gains from reclamation, revegetation and biodiversity offsets ultimately exceed impacts to native vegetation within the combined mine footprints. Methods Study site The study area encompassed the combined past and current footprints of all Teck steelmaking coal mining operations within the Elk River Valley (“Elk Valley”) of southeast British Columbia. The Elk Valley is located along the western edge of the Rocky Mountains, in terrain dominated by steep slopes and narrow valley bottoms. The valleys are generally underlain by sandy and shale rocks of the Jurassic and Cretaceous eras. Well-drained soils usually occur on steep slopes with a thin layer of surficial material over bedrock. Areas with poor drainage occur on lower and toe slope positions and in isolated areas along nearly level terrain. The Elk Valley is largely forested, mainly of coniferous stands. The most common Biogeoclimatic zones include the Engelmann Spruce–Subalpine Fir Dry Cool variant and Engelmann Spruce–Subalpine Fir Dry Cool Woodland subzone (Meidinger and Pojar 1991 ). Teck’s steelmaking coal operations within the Elk River Valley include four active mines and one under care and maintenance and have a combined area of c. 19,000 ha. This total includes areas currently disturbed by mining plus areas that were disturbed in the past but are now reclaimed or undergoing reclamation, through 2018. Vegetation mapping and classification For the baseline assessment, vegetation of the mine footprint was reconstructed using a combination of archival aerial photos, satellite imagery, and Predictive Ecosystem Mapping (PEM; Terrestrial Ecosystem Mapping Alternatives Task Force Resource Inventory Committee (RIC) 1999 ) as well as terrain reconstruction and hydrological and soil models. The resulting back-cast vegetation polygons were combined with maps of existing vegetation to produce a comprehensive baseline map of pre-mine vegetation conditions within all Elk Valley operations administered by Teck (Knopff and Franklin 2018 ). All vegetation units were classified to the Site Series level, using the Biogeoclimatic Ecosystem Classification (BEC) version 10 (British Columbia Ministry of Forests-Lands and Natural Resource Operations 2016 ). Seral stage classes were assigned to forested vegetation using BC Ministry of Environment and Climate Change Strategy standard terminology (BC Ministry of the Environment 2010 ). As detailed information on vegetation condition at the baseline year was not available, we used the condition of current vegetation outside the mine footprint to represent vegetation condition within the mine footprints at baseline. We believe this assumption is conservative as land management practices have generally improved over time and anthropogenic fire along with other large-scale land disturbance practices have decreased; therefore, current condition should be as good as, or better than, condition at the baseline year (although we acknowledge that fire may have contributed to increased quality for drier forest types which were structured by fire in pre-European contact times). For much of the 20th century, the extent of old growth forest within the Elk Valley was considerably less than under historical disturbance regimes prior to European settlement in the mid-1850s. In particular, the lack of old growth Montane Spruce (MS) and Interior Cedar Hemlock (ICH) forest reflects a combination of very large and severe fires in 1919 and in the 1930s, forest clearing for urban/rural development and agriculture, and timber harvesting that has been ongoing since the early 1900s (Holmes et al. 2018 ). As we did not subdivide native vegetation into subcategories of disturbance, each vegetation type is represented by exactly one focal land cover class only. The resulting vegetation and land cover map for the study site consisted of 133 site series (MacKenzie 2012 ; British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) 2016 ). Including structural stages of forested vegetation, the final number of distinct land cover classes (sites series + structural stages) was 396. As this number of sampling strata was too large to be sampled at adequate sample sizes, we revised the baseline map, combining site series into a smaller number of classes (hereafter referred to as "ecosystem groups") representing vegetation of similar structure, composition and moisture availability. For example, site series representing above-treeline alpine vegetation were combined into the ecosystem group “Alpine”. In addition, for forested vegetation, the 14 BEC structural stages were combined into the three seral stages “Early-mid”, “Mature” and “Old”. After these simplifications, the final vegetation and land cover map consisted of 21 ecosystem groups plus seral stage sampling strata (Table 1 ). Five sampling strata (Alpine grassland, Alpine meadow, Deciduous floodplain, Herb meadow and Shrubland) had focal or benchmark sample sizes that were too low (0–4 plots) to allow analysis. The total area of all units within the footprint study area was 18,978.9 ha. Table 1 Benchmark and focal plot sample sizes for all vegetation classes (ecosystem groups) within the study area, for the baseline and current (reclamation) assessments. Sample sizes will be smaller than shown for individual indicators not measured in all plots. Focal plots Ecosystem group Benchmark plots Baseline native vegetation Current reclaimed vegetation Alpine 10 81 Alpine Dwarf Shrub 46 30 Alpine grassland 4 41 Alpine meadow 2 15 Avalanche feature 100 24 Brushland/Grassland 62 61 98 Deciduous floodplain 0 4 Dry forest, Early-mid 79 95 105 Dry forest, Mature 55 70 Dry forest, Old 14 4 Herb meadow 0 6 Intermediate forest, Early-mid 148 220 10 Intermediate forest, Mature 83 216 Intermediate forest, Old 28 22 Krummholz 15 9 Rock/Talus 64 60 Shrubland 2 6 Wet forest, Early-mid 70 46 9 Wet forest, Mature 77 54 Wet forest, Old 28 11 Wetland 72 46 For the current vegetation assessment, all reclaimed vegetation types within the footprint were assigned a target benchmark vegetation and the total area of each reclamation class was determined. Reclaimed vegetation fell into four ecosystem groups: “Brushland/Grassland”, “Early-mid Dry Forest”, “Early-mid Intermediate Forest” and “Early-mid Wet Forest”. As our accounting encompassed past, present and future impacts through 2035, including planned expansions, vegetated areas currently falling within the footprint of future mine expansions were coded as "non-vegetated" and included with the un-reclaimed areas within the current mine footprint. Total area of reclaimed vegetation was 2,109.0 ha, with the remaining non-vegetated areas totaling 16,869.9 ha. Vegetation plots The baseline vegetation plots were sampled over a 41-year period post-dating the baseline year, from 1975–2016, in native vegetation adjacent to the project site (i.e., the current mine footprint). These samples provide a conservative estimate of the condition of native vegetation present prior to mining impacts (see "Vegetation mapping and classification"). These data were collected by multiple consulting companies as part of Teck’s ongoing vegetation mapping and inventory program within the Elk Valley. After removing plots not compatible with BC ENV Site Visit methodology (BC Ministry of the Environment 2010 ) the final baseline focal dataset consisted of 1,121 vegetation plots. Vegetation plots for the current assessment were sampled in reclaimed vegetation within the project footprint. These plots are located in multiple areas representing different reclamation treatments. For many of them, reclamation was initiated more than a decade ago. Unlike current practices, many historical reclamation activities did not have restoration of native vegetation as a primary goal, and included planting of non-native agronomic species, particularly in grassland ecosystems targeted to support an ungulate end land use. Despite potential differences in reclamation methods and initiation times, we did not distinguish among different reclamation treatments. Our goal for this preliminary VQA application was to gain a general understanding of the current state of historical reclamation for broad classes of vegetation. This information will be used to plan more fine-grained sampling in future assessments, and enable adaptation of reclamation practices to better serve Teck's current focus on native vegetation. All current assessment focal plots were collected using the same Site Visit methodology as the baseline assessment plots. The final reclamation data set consisted of 222 plots. Benchmark plots were obtained from a database of inventories developed by the BC ENV as reference data for the BEC vegetation classification standard (Meidinger and Pojar 1991 ; MacKenzie 2012 ; British Columbia Ministry of Forests-Lands and Natural Resource Operations 2016 ). These data were also recorded using the BC ENV Site Visit methodology (BC Ministry of the Environment 2010 ). After selecting the subset of inventories which matched at the Site Series level to native vegetation units present within the focal site baseline map (see "Vegetation mapping and classification"), the final benchmark dataset consisted of 959 vegetation plots. Table 1 shows the distribution of the benchmark, baseline and reclamation vegetation plots among the 21 ecosystem groups. Ecological Indicators We selected ten attributes that were recorded with sufficient consistency across all data sets to be used as ecological indicators (Table 2 ). Three of these (“Species richness”, "Taxonomic distance" and “Percent cover exotic species”) were taxonomic-based indicators which we combined within functional group "Composition". Four (“Percent cover herbs”, “Percent cover moss”, “Percent cover shrubs”, “Percent cover trees”) reflected aspects of vegetation height and complexity and were therefore placed in functional group "Structure". The remaining three (“Percent cover dead wood”, “Percent cover organic soil” and “Percent cover surface water”) were aspects of the physical environment which influence vegetation composition and structure; these were assigned to functional group "Environment". Indicator “Percent cover surface water” was included only in quality calculations for wetland vegetation (ecosystem group “Wetland”). “Species richness” was the count of all native species recorded in the plot. Native and exotic (introduced) species status was assigned using the 2016 edition of the BC Flora Checklist (BC Ministry of the Environment 2016 ). Indicator “Taxonomic distance” (TD) measured dissimilarity of species composition of the focal plots relative to the benchmark plots, as determined using non-metric multidimensional scaling (NMDS); details of the calculation of this indicator are provided in Text S6 and Fig. S5. The remaining indicators were measured as described in BC Ministry of the Environment ( 2010 ). Table 2 Ecological indicators used for the example VQA application, their probability distributions and quality scoring methods. Indicator Distribution Quality algorithm Test tail Functional group Species Richness Negative binomial Overlap Lower Composition Taxonomic Distance Gamma Overlap Both Composition Percent Cover Exotic Species Beta Fixed (0) n/a Composition Percent Cover Herbs Beta Overlap-means Both Structure Percent Cover Moss Beta Overlap-means Both Structure Percent Cover Shrubs Beta Overlap-means Both Structure Percent Cover Trees Beta Overlap-means Both Structure Percent Cover Decomposing Wood Beta Overlap-means Both Environment Percent Cover Organic Soil Beta Overlap-means Both Environment Percent Cover Surface Water* Beta Overlap-means Both Environment *Indicator “Percent cover surface water” used for “Wetland” vegetation only. Quality and quality hectares We calculated indicator quality and overall quality for sampling strata with at least 6 focal and 6 benchmark plots. Quality of “Percent cover exotic species” was calculated using a fixed benchmark value of 0. All other indicators used the overlap or overlap-means quality algorithms (Table 2 ). The 95% CLs of quality were calculated using 10,000 bootstrap iterations. QH and QH.net were calculated following the procedure described in "Loss-gain accounting". Vegetation types without at least 6 focal and 6 benchmark plots were excluded from quality calculations. For these vegetation types, we conservatively assumed a baseline quality of 100% and therefore used their actual areas as Quality Hectares in all subsequent calculations. Effect size, power, and sample size We performed Monte Carlo simulations for each vegetation type of the relationship between sample size ( n ; equal for focal and benchmark) and power to detect a significant change in overall quality at α = 0.05 and an effect size (MDES) of 0.1, as described in "Effect size, power, and minimum sample size". For each class of vegetation, the required minimum sample size, n min , was the value of n at which a logistic fit to a plot of n and power for all simulations crossed the conventional power line of 0.80. We also performed power-sample size simulations for selected indicator-vegetation combinations. As some indicators required very large minimum sample sizes to achieve an MDES of 0.1, we also ran simulations at MDESs of 0.15 and 0.2 to examine the trade-off between minimum sample size and sensitivity. Software All VQA calculations were implemented in the R programming language (R Development Core Team 2008). Univariate pdfs were fit to benchmark and focal samples by maximum likelihood estimation using function fitdistr of package MASS (Ripley et al. 2013 ). Results and discussion Baseline assessment Baseline overall quality [lower, upper 95% CLs] of the assessed vegetation types ranged from a low of 0.74 [0.64, 0.88] for Krummholz to a high of 0.87 [0.83–0.95] for Intermediate forest, Mature (Table 3 ; SM-1). The below-benchmark overall quality scores of forested vegetation, plus details of individual indicators (see below) were consistent with a history of logging in the early 1900s, prior to the initiation of the mines. Distributions of individual indicators and their quality scores (Table S1 ) support this interpretation. For early-mid seral stages of all three forest types, a disproportionate number of focal sites had little to no large tree cover and consequently low quality scores for indicator Percent Cover Trees (Fig. 3 a for Intermediate Forest, Early-mid; results for Early-mid Dry Forest and Wet Forest similar). These anomalies were structural only; taxonomic composition of early-mid seral forests was almost identical to benchmark (e.g., Fig. 3 b). In addition, strongly bimodal focal distributions of indicator Percent Cover Trees suggest that the Early-mid seral forest class in the Elk Valley may represent mixtures of naturally regenerating forest and recovering forest harvest. If true, the accuracy of the baseline assessment would be improved by splitting each stratum into two disturbance classes, one natural and one disturbed, sharing the same benchmark vegetation, and assessing each class independently with a separate set of focal plots. This conclusion highlights the value of examining individual indicator distributions and quality scores in addition to overall quality. Table 3 Baseline and current (reclaimed) vegetation quality for the combined Teck Elk Valley Operations footprint. For vegetation for which quality could not be calculated due to insufficient sample size (indicated by *), 100% quality is assumed. Total footprint area (all ecosystem groups combined) is 18,978.9 ha. CLs: confidence limits. Ecosystem group Baseline quality [95% CLs] Reclaimed (current) quality [95% CLs] Alpine 0.79 [0.72–0.88] 0 Alpine dwarf shrub 0.76 [0.68–0.85] 0 Alpine grassland* 1 0 Alpine meadow* 1 0 Avalanche feature 0.86 [0.81–0.95] 0 Brushland/Grassland 0.82 [0.78–0.88] 0.35 [0.32–0.40] Deciduous floodplain* 1 0 Dry forest, early-mid 0.81 [0.76–0.90] 0.42 [0.39–0.46] Dry forest, mature 0.84 [0.79–0.94] 0 Dry forest, old* 1 0 Herb meadow* 1 0 Intermediate forest, early-mid 0.79 [0.76–0.91] 0.31 [0.28–0.36] Intermediate forest, mature 0.87 [0.83–0.95] 0 Intermediate forest, old 0.76 [0.69–0.83] 0 Krummholz 0.74 [0.64–0.88] 0 Rock/Talus 0.82 [0.77–0.89] 0 Shrubland* 1 0 Wet forest, early-mid 0.75 [0.70–0.83] 0.26 [0.19–0.32] Wet forest, mature 0.81 [0.75–0.90] 0 Wet forest, old 0.82 [0.74–0.90] 0 Wetland 0.83 [0.78–0.90] 0 Non-forested vegetation classes also had overall quality scores < 100% (i.e., CIs not overlapping 100%); thus, factors other than forest harvesting were responsible for the below-benchmark baseline vegetation quality. In the case of non-forested alpine vegetation (Alpine, Alpine Dwarf Shrub, Avalanche Feature, Krummholz), the somewhat lower quality scores for structural and environmental indicators (Table S2) are consistent with the moderate structural degradation noted by previous studies and attributed to intense grazing by native ungulates (eg, Poole et al., 2020 ). At lower elevations, quality of indicator "Percent Cover Exotic Species” for Brushland/Grassland (0.77 [0.66, 0.85]) was lower than other forested and non-forested vegetation type (all of which had CIs close to or overlapping 100%; see Table S1 ). Although the majority of plots in Brushland/Grassland had zero cover of exotic species, a few had high abundance, with exotic cover at some focal sites exceeding 60% (Fig. 3 c). A number of benchmark plots (not used to calculate this indicator; see Table 2 ) also had exotic species cover as high as ~ 30%. The prevalence of exotic species in Brushland/Grassland is also reflected in its somewhat anomalous species composition and low score for Taxonomic Distance (Fig. 3 d). These observations are consistent with a general pattern of invasion of western North American grasslands and shrublands by exotic species (Dukes and Mooney 2004 ). Reclamation assessment Of the original 18,978.9 ha within the study area mine footprint, 2,109.0 ha had been re-vegetated at time of this analysis to the four reclamation vegetation classes shown in Table 3 . Overall quality of reclaimed vegetation ranged from 0.26 [0.19–0.32] for Wet forest, Early-mid to 0.42 [0.39–0.46] for Dry Forest, Early-mid, with no upper CLs exceeding 0.46. Quality of the individual indicators paints a clearer picture of the ways in which each type of historical reclamation differed from undisturbed vegetation. Brushland/Grassland had fewer species, different taxonomic composition and much higher cover of non-native species, relative to the benchmark condition (Fig. 4 ; see also Table S3). These results are consistent with the widespread use of a small number of non-native agronomic species in historical grassland reclamation, which was guided largely by provincially-approved standards for ungulate forage production. Reclaimed forest differed strongly from naturally occurring early-mid seral forest in both structure (Fig. 5 ) and composition (Tables S3 & S4), with anomalously low cover of mosses, shrubs and trees (Fig. 5 b-d), lower species richness and different taxonomic composition (Fig. S6). Sample size requirements Simulations of power versus sample size indicated that sample sizes ranging from 9–17 were required to detect a change in overall quality of 0.1 in the baseline vegetation (e.g., Fig. 12). By contrast, much smaller sample sizes were needed to detect a similar effect size in the reclaimed vegetation (Fig. 11a). This difference reflects the much narrower CIs of overall quality in the reclaimed vegetation (Table 3 ), which in turn are a consequence of the low variance of many reclaimed vegetation indicators (i.e., Figs. 9b, 9c, 9d and 10a; note especially the predominance of values close to or exactly equal to zero). The low minimum sample sizes shown in Fig. 6 b should not be taken literally; considerations of power and effect size aside; sample sizes as small as n = 3 are unlikely to be representative of an entire class of reclamation. Furthermore, as reclaimed vegetation matures and the variances of its indicators and overall quality converge upon values observed in the benchmark vegetation, n.min should also increase. Ultimately, minimum sample sizes for mature reclaimed vegetation will be similar to the values of n.min determined for the baseline vegetation (e.g., Fig. 6 a). In practice, smaller samples of n = 6 plots can be installed initially for monitoring reclamation quality during the first few years, and additional plots added incrementally as the vegetation matures. Sample size requirements for individual indicators were generally higher than for overall quality. For example, samples sizes of 18, 27 and 42 were required to achieve effect sizes of 0.2, 0.15 and 0.1, respectively, for Species Richness of vegetation "Avalanche Feature", (Fig. S7). Given the importance of the indicator distributions and indicator quality for diagnosing and resolving causes of low overall quality scores, we recommend larger sample sizes than the minimum required for the measurement of overall quality. Finally, we must emphasize that simulation-based estimates of minimum samples sizes are approximations only, and should be evaluated in light of the actual CLs obtained and by repeating power simulations as more data become available. If an acceptable level of error is attained prior to reaching the estimated minimum sample size, no additional sampling is needed. An even more compelling reason for targeting higher minimum sample sizes than those suggested here is to minimize the amount by which the estimated current QH must exceed baseline QH to ensure that the lower 95% CL of Net QH is greater than zero (see Fig. 1 b). The decision to target a more stringent MDES of 0.05 or lower by increasing sample sizes can be made on a case by cases basis by weighing the costs of increased sampling versus additional restoration and offset obligations. Quality Hectares and progress to NPI Comparison of current to baseline QHs shows the contribution of historical reclamation to the overall goal of NPI (Table 5 ). Although impacts to most classes of vegetation will need to be compensated via offset actions elsewhere, on-site restoration has made major contributions toward NPI for Brushland/Grassland and Early-mid Dry Forest; 62% and 45%, respectively, of the baseline QH of these two vegetation types had been restored by the time these data were collected. In addition to extending re-vegetation efforts to areas not currently undergoing reclamation, additional gains in QH could be achieved by improving the quality of existing and future reclamation. This is particularly true of Brushland/Grassland, where replacement of non-native grasses with native species has the potential to provide substantial improvements in quality for multiple indicators (see "Reclamation assessment"). In the case of Early-mid Dry Forest, this vegetation is expected to transition over time to Mature Dry Forest, and restoration should therefore aim to restore substantially more than the number of QHs originally present of the early-mid seral stage. Over-planting of the earliest seral stages will ensure a more balanced restoration of seral stages at and beyond closure. Teck has recently refocused the current reclamation program to target native species as well as aiming towards natural ecosystem assemblages for end land use objectives (Franklin and Burton 2018 ). In this context, VQA provides a mechanism for integrating the contribution of native species and natural ecosystem assemblages into the goal of NPI. Finally, it is important to emphasize that CLs, not means, determine when NPI has been attained. As illustrated in Fig. 1 b, only when the lower 95% CL of Net QH exceeds zero (QH.net.lcl > 0) are we justified in concluding that our estimate of QH.net also exceeds zero. It should also be apparent that excessively broad CIs can increase restoration obligations beyond what would be required given a more accurate estimate of QH.net. The incorporation of measurement error into the VQA workflow thus creates a built-in incentive for estimating quality, QH, and net QH with the greatest possible accuracy, subject to the opposing constraint of increased sampling and monitoring costs. Table 5 Baseline area and Quality Hectares (QH), current (reclaimed) area and QH, and net QH of vegetation within the combined Teck Elk Valley Operations footprint. Baseline and current quality hectares are the product of the areas shown here and the quality scores in Table 3 . Net QH is current QH - baseline QH. For vegetation for which quality could not be calculated due to low sample size (indicated by asterisk), baseline QH = actual area (ha). CLs: confidence limits. Vegetation (ecosystem group) Baseline Current (reclaimed) Net QH [95% CLs] Area (ha) Quality Hectares [95% CLs] Area (ha) Quality Hectares [95% CLs] Alpine 385.0 304.2 [277.2, 338.8] 0 0 -304.2 [-338.8, -277.2] Alpine dwarf shrub 0.6 0.5 [0.41, 0.51] 0 0 -0.5 [-0.5, -0.4] Alpine grassland* 67.1 67.1 0 0 -67.1 Alpine meadow* 13.0 13.0 0 0 -13 Avalanche feature 741.5 637.7 [600.6, 704.4] 0 0 -637.7 [-704.4, -600.6] Brushland/Grassland 862.0 706.8 [672.4, 758.6] 1257.7 440.2 [402.5, 503.1] -266.6 [-269.9, -255.5] Deciduous floodplain* 157.3 157.3 0 0 -157.3 Dry forest, early-mid 920.4 745.5 [699.5, 828.4] 800.8 336.3 [312.3, 368.4] -409.2 [-460, -387.2] Dry forest, mature 521.4 438.0 [411.9, 490.1] 0 0 -438 [-490.1, -411.9] Dry forest, old* 94.1 94.1 0 0 -94.1 Herb meadow* 0.3 0.3 0 0 -0.3 Intermediate forest, early-mid 7,771.00 6139.1 [5,906.0, 7,071.6] 3.5 1.1 [1.0, 1.3] -6,138 [-7,070.3, -5,905] Intermediate forest, mature 3,931.00 3420 [3,262.7, 3,734.5] 0 0 -3,420 [-3,734.5, -3,262.7] Intermediate forest, old 1,272.00 966.7 [877.7, 1,055.8] 0 0 -966.7 [-1,055.8, -877.7] Krummholz 55.6 41.1 [35.6, 48.9] 0 0 -41.1 [-48.9, -35.6] Rock/Talus 233.4 191.4 [179.7, 207.7] 0 0 -191.4 [-207.7, -179.7] Shrubland* 9.4 9.4 0 0 -9.4 Wet forest, early-mid 923.0 692.3 [646.1, 766.1] 47.0 12.2 [8.9, 15.0] -680.1 [-751.1, -637.2] Wet forest, mature 565.0 457.7 [423.8, 508.5] 0 0 -457.7 [-508.5, -423.8] Wet forest, old 317.0 259.9 [234.6, 285.3] 0 0 -259.9 [-285.3, -234.6] Wetland 138.8 115.2 [108.3, 124.9] 0 0 -115.2 [-124.9, -108.3] Conclusions The VQA approach provides a quantitative, objective framework for measuring the extent and quality of pre-impact vegetation, determining compensation obligations, monitoring the pace and effectiveness of restoration, and determining the point at which NPI has been achieved. By using distribution overlap as the main measure of quality, the method incorporates natural, landscape-scale variation not captured by other methods. Examination of the distributions of individual indicators provides insights that can be used to fine-tune restoration practices and adjust offset goals. Random sampling reduces vulnerability to bias and enables statistical inference at the scale of entire landscape. The ability to quantify uncertainty increases confidence in monitoring results and creates built-in incentives for maximizing accuracy of quality estimates. A potential shortcoming of the method is the need for relatively large sample sizes. This shortcoming is balanced in part by the relatively small number of indicators required, and the use of sampling methods familiar to most field crews experienced in vegetation surveys. In addition, the ability to use existing vegetation data can reduce the number of new plots needed. Over time, the growth of repositories of vegetation inventory data will lower the number of new benchmark plots that need to be collected. Regulators should encourage companies to deposit samples of benchmark-quality vegetation in publicly available databases; companies themselves may voluntarily contribute such data in the interest of reducing future sampling burdens. Declarations Acknowledgments. We thank Jared Hardner, Garry Luini, Katrina Lukianchuk, and Justin Straker and for suggestions that helped improve aspects of the VQA framework, and Christina Small for comments that strengthened the final manuscript. Dan Vasiga prepared the Elk Valley database exports of plots, vegetation and land cover used for the example application. Will Mackenzie and Deb MacKillop (BCENV) assisted with identification and selection of benchmark-quality plot data from the BCENV provincial vegetation database. This work was supported in part by funding from Teck Coal Limited. References BC Ministry of the Environment (2021) Terrestrial Ecosystem Data Standards & Guidelines. https://www2.gov.bc.ca/gov/content/environment/plants-animals-ecosystems/ecosystems/tei-standards BC Ministry of the Environment (2010) Field Manual for Describing Terrestrial Ecosystems, 2nd Editio. B.C. Ministry of Forests and Range, B.C. Ministry of Environment BC Ministry of the Environment (2016) BC Flora checklist 2016. https://www.for.gov.bc.ca/hre/becweb/resources/codes-standards/standards-species.html . Accessed 18 Jul 2016 Berger VW, Zhou Y (2014) Kolmogorov–Smirnov Test: Overview. In: Wiley StatsRef: Statistics Reference Online. John Wiley & Sons, Ltd Best DJ, Gipps PG (1974) An Improved Gamma Approximation to the Negative Binomial. Technometrics 16:621–624 British Columbia Ministry of Forests-Lands and Natural Resource Operations (2016) BECMaster ecosystem plot database [MSAccess 2003 format]. Research Branch, Victoria, B.C. http://www.for.gov.bc.ca/hre/becweb/resources/information-requests/index.html British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) (2016) Biogeoclimatic Ecosystem Classification, Version 10 CBD (2020) Update of the zero draft of the post-2020 global biodiversity framework Ceballos G, Ehrlich PR, Barnosky AD, et al (2015) Accelerated modern human – induced species losses: entering the sixth mass extinction. Sci Adv 1:1–5. https://doi.org/10.1126/sciadv.1400253 Chazdon RL, Letcher SG, van Breugel M, et al (2007) Rates of change in tree communities of secondary Neotropical forests following major disturbances. Philos Trans R Soc Lond B Biol Sci 362:273–89. https://doi.org/10.1098/rstb.2006.1990 Cook CN, Wardell-Johnson G, Keatley M, et al (2010) Is what you see what you get? Visual vs. measured assessments of vegetation condition. J Appl Ecol 47:650–661. https://doi.org/10.1111/j.1365-2664.2010.01803.x Cousineau D (2017) Varieties of confidence intervals. Adv Cogn Psychol 13:140–155. https://doi.org/10.5709/acp-0214-z Cumming G (2014) The New Statistics: Why and How. Psychol Sci 25:7–29. https://doi.org/10.1177/0956797613504966 Curtin CG (1995) Can montane landscapes recover from human disturbance? Long-term evidence from disturbed subalpine communities. Biol Conserv 74:49–55. https://doi.org/10.1016/0006-3207(95)00014-U Demeo T, Haugo R, Ringo C, et al (2018) Expanding Our Understanding of Forest Structural Restoration Needs in the Pacific Northwest. Northwest Sci 92:18–35. https://doi.org/10.3955/046.092.0104 Dukes JS, Mooney HA (2004) Disruption of ecosystem processes in western North America by invasive species. Rev Chil Hist Nat 77:411–437 Eddy D, Hall R, Rehwinkel R, et al (2011) Assessing the assessors: Quantifying observer variation in vegetation and habitat assessment. 12:144–148 Efron B (1979) Bootstrap Methods: Another Look at the Jackknife. Ann Stat 7:1–26. https://doi.org/10.1214/aos/1176344552 Eyre TJ, Kelly A., Neldner VJ, et al (2011) BioCondition: A Condition Assessment Framework for Terrestrial Biodiversity in Queensland. Assessment Manual. Version 2.1. Brisbane Franklin CW, Burton A (2018) End land use objective planning : integrating an ecosystem based approach into biodiversity and reclamation planning. In: 41st Annual TRCR Mine Reclamation Symposium. Fort Williams, BC, Canada, pp 1–11 Fulé PZ, Covington WW, Moore (1997) Determining reference conditions for ecosystem management of southwestern ponderosa pine forests. Ecol Appl 7:895–908 Gastwirth JL (1975) Statistical measures of earnings differentials. Am Stat 29:32–35. https://doi.org/10.1080/00031305.1975.10479109 Gentry AH (1993) Patterns of diversity and floristic composition in neotropical montane forests. In: Churchill SP, Balslev H, Forero E, Luteyn J (eds) Proceedings of the Neotropical Montane Forest Biodiversity and Conservation Symposium, The New York Botanical Garden, 21–26 June 1993. New York Botanical Garden, Bronx, N.Y., pp 103–126 Gentry AH (1988) Changes in plant community diversity and floristic composition on environmental and geographical gradients. Ann Missouri Bot Gard 75:1–34 Gentry AH (1995) Diversity and floristic composition of neotropical dry forests. In: Bullock SH, Mooney HA, Medina E (eds) Seasonally Dry Tropical Forests. Cambridge University Press, Cambridge, pp 146–194 Gibbons P, Ayers D, Seddon J, et al (2008) Biometric 2.0: A Terrestrial Biodiversity Assessment Tool for the NSW Native Vegetation Assessment Tool - Operational Manual. Canberra Gibbons P, Freudenberger D (2006) An overview of methods used to assess vegetation condition at the scale of the site. Ecol Manag Restor 7:S10–S17. https://doi.org/10.1111/j.1442-8903.2006.00286.x Holmes P, Stuart-Smith K, MacKillop D, et al (2018) Old and Mature Forest Cumulative Effects Assessment Report: Elk Valley, Kootenay-Boundary Region Hurlbert SH (1984) Pseudoreplication and the design of ecological experiments. Ecol Monogr 54:187–211 International Finance Corporation (2012) IFC Sustainability Framework: Policy and performance standards on environmental and social sustainability. Washington, DC, USA Jennings MD, Faber-Langendoen D, Loucks OL, et al (2009) Standards for associations and alliances of the U.S. National Vegetation Classification. Ecol Monogr 79:173–199. https://doi.org/10.1890/07-1804.1 Kappelle M (1995) Ecology of mature and recovering Talamancan montane Quercus forests, Costa Rica. University of Amsterdam, Amsterdam Kier G, Dinerstein E, Ricketts TH, et al (2005) Global patterns of plant diversity and floristic knowledge. J Biogeogr 32:1107–1116 Knopff K, Franklin WE (2018) Biodiversity management: establishing pre-existing baseline conditions on mature and historical mining disturbances to derive back-casted wildlife habitat suitability model metrics for ten wildlife species. In: 41st Annual TRCR Mine Reclamation Symposium. Fort Williams, BC, Canada Korb JE, Covington WW, Fulé PZ (2003) Sampling Techniques Influence Understory Plant Trajectories After Restoration : An Example from Ponderosa Pine Restoration. 11:504–515 Landres PB, Morgan P, Swanson FJ, et al (2014) Overview of the Use of Natural Variability Concepts in Managing Ecological Systems. Ecol Appl 9:1179–1188 Lewis M, Christianson A, Spinks M (2018) Return to flame: Reasons for burning in Lytton First Nation, British Columbia. J For 116:143–150. https://doi.org/10.1093/jofore/fvx007 MacKenzie W (2012) Biogeoclimatic ecosystem classification of non-forested ecosystems in British Columbia. Victoria, B.C. Magurran AE (2004) Measuring biological diversity. Blackwell Publishing Ltd, Oxford, UK Maron M, Hobbs RJ, Moilanen A, et al (2012) Faustian bargains? Restoration realities in the context of biodiversity offset policies. Biol Conserv 155:141–148. https://doi.org/10.1016/j.biocon.2012.06.003 McCarthy M, Parris K (2004) The habitat hectares approach to vegetation assessment: An evaluation and suggestions for improvement. Ecol Manag Restor 5:24–27 McElhinny C, Gibbons P, Brack C, Bauhus J (2005) Forest and woodland stand structural complexity: Its definition and measurement. For Ecol Manage 218:1–24. https://doi.org/10.1016/j.foreco.2005.08.034 Meidinger D V, Pojar J (1991) Ecosystems of British Columbia Parkes D, Newell G (2003) Assessing the quality of native vegetation: the “habitat hectares” approach. Ecol Manag Restor 4:29–38 Peet RK, Lee MT, Jennings MD, D. Faber-Langendoen D (2012) VegBank: a permanent, open-access archive for vegetation plot data. Biodivers Ecol 4:233–241 Pimm SL, Joppa LN (2015) How Many Plant Species are There, Where are They, and at What Rate are They Going Extinct? Ann Missouri Bot Gard 100:170–176. https://doi.org/10.3417/2012018 Poole K, Teske I, Podrasky K, et al (2020) Bighorn Sheep Cumulative Effects Assessment Report: Elk Valley, Kootenay-Boundary Region Rainey HJ, Pollard EHB, Dutson G, et al (2015) A review of corporate goals of No Net Loss and Net Positive Impact on biodiversity. Oryx 49:232–238. https://doi.org/10.1017/S0030605313001476 Rio Tinto (2008) Rio Tinto and biodiversity: achieving results on the ground. Rio Tinto plc and Rio Tinto Limited Ripley B, Venables B, Bates DM, et al (2013) Package ‘MASS’ Sahley CT, Vildoso B, Casaretto C, et al (2017) Quantifying impact reduction due to avoidance, minimization and restoration for a natural gas pipeline in the Peruvian Andes. Environ Impact Assess Rev 66:53–65. https://doi.org/10.1016/j.eiar.2017.06.003 Schlesinger W, Andrews J (2000) Soil respiration and the global carbon cycle. Biogeochemistry 48(1):7–20 Terrestrial Ecosystem Mapping Alternatives Task Force Resource Inventory Committee (RIC) (1999) Towards the Establishment of Predictive Ecosystem Mapping Standards : A White Paper The State of Queensland (2014) BioCondition Benchmarks. https://www.qld.gov.au/environment/plants-animals/biodiversity/benchmarks/ Venables WN, Ripley BD (2002) Modern Applied Statistics with S, 4th ed. Springer Science & Business Media Additional Declarations Competing interest reported. BLB was funded in part by Teck during the preparation of this manuscript. Some of the paid consulting work performed by BLB and REG involves application of the methods described in the manuscript. AB and WF are employees of Teck. Supplementary Files BoyleetalSupplementaryMaterials.docx Cite Share Download PDF Status: Published Journal Publication published 01 Jan, 2024 Read the published version in Ecological Indicators → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3374961","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":235309544,"identity":"813c532d-45cd-4ca7-9dd6-820df20c6b31","order_by":0,"name":"Bradley L. Boyle","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAz0lEQVRIiWNgGAWjYBACAwkg8YDBhoGBGSZ0gBgtCQxppGs5jCRESIu5dPMxiYSK8/Lm7ewPHxfUMMjx3UjAr8VyzrE0iYQztw3nHGZINp5xjMFYkpAWgxs5ZhKJbbcZZzAzHJPmYWNI3ECcln/n7GcwM7ZJ8/xjqCdSS8OBxBnMzGzSvG0MCQZE+CXZIuFYcvIMZjZmY94+CcOZZx7g1wIMsYM3PtTY2c7gP/7wMc83G3m+4wRsAQIWCSSOBE5lyID5A1HKRsEoGAWjYOQCAKMkQhY23gr9AAAAAElFTkSuQmCC","orcid":"","institution":"University of Arizona","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Bradley","middleName":"L.","lastName":"Boyle","suffix":""},{"id":235309545,"identity":"7f101fa9-a3d6-46d7-9be9-c1aba84fc12c","order_by":1,"name":"Warn Franklin","email":"","orcid":"","institution":"Teck Coal Limited","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Warn","middleName":"","lastName":"Franklin","suffix":""},{"id":235309546,"identity":"5add6cc5-5d18-446f-85e5-5266dd3f2c3f","order_by":2,"name":"Alison Burton","email":"","orcid":"","institution":"Teck Coal Limited","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Alison","middleName":"","lastName":"Burton","suffix":""},{"id":235309548,"identity":"ed5c60ce-1211-4153-8dd9-1d28a76bfcde","order_by":3,"name":"Raymond E. Gullison","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Raymond","middleName":"E.","lastName":"Gullison","suffix":""}],"badges":[],"createdAt":"2023-09-21 07:14:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3374961/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3374961/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1016/j.ecolind.2023.111510","type":"published","date":"2024-01-01T06:18:08+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":43843636,"identity":"384f17ad-9dbe-4c67-afa9-a3d98fd32377","added_by":"auto","created_at":"2023-09-28 15:50:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":84731,"visible":true,"origin":"","legend":"\u003cp\u003eThe VQA workflow. (a) Change in Quality Hectares (QH) over time for a single vegetation type is measured by a series of QH assessments, beginning with a baseline assessment (Baseline QH) prior to any project impacts. Additional QH from offset sites (upper line) are added to QH at the project site to help achieve a Net Positive Impact (NPI). (b) Losses and gains expressed as Net QH (current QH - baseline QH, or QH\u003csub\u003enet\u003c/sub\u003e), with 95% confidence intervals (CI) indicated by vertical bars. Baseline QH\u003csub\u003enet\u003c/sub\u003e is, by definition, zero, and NPI is achieved when the lower 95% confidence limit (CL) of QH\u003csub\u003enet\u003c/sub\u003e exceeds zero.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/71758a5a265fc8a3e9cd23d4.png"},{"id":43843640,"identity":"8600be7e-cbc6-4b0b-bfea-1ac6037df252","added_by":"auto","created_at":"2023-09-28 15:50:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":102161,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eCalculation of overlap-based quality. In (a) and (b), separate probability density functions (pdfs) are fit to the focal and benchmark sampling distributions, respectively, using maximum likelihood. In (c), quality [95% CLs in square brackets] is calculated as the overlap of the two pdfs. Dashed lines in (c) are the bootstrapped estimates of the 95% CLs of the pdfs; pale gray lines are a small sample of the individual bootstrap pdfs. CLs: confidence limits.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/1a3b0fe2fffc8c28be6e782d.png"},{"id":43843637,"identity":"1c0feb30-be93-4521-aae6-f154351953ee","added_by":"auto","created_at":"2023-09-28 15:50:19","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":106566,"visible":true,"origin":"","legend":"\u003cp\u003eObserved (histograms) and fitted (curves) focal and benchmark distributions of selected indicators for baseline Intermediate Forest, Early-mid (top panels) and Brushland/Grassland (bottom panels), and their quality scores [lower, upper 95% CLs]. Vertical dashed line in (c) indicates the fixed benchmark value of 0 used for this indicator. Arrow in (c) marks the focal mean of 0.77. Pink: focal; blue: benchmark; purple: overlap.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/086f250429cabe78329b5a43.png"},{"id":43843664,"identity":"212a6535-c1d5-4890-83ad-6e600ff06cb2","added_by":"auto","created_at":"2023-09-28 15:50:20","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":70191,"visible":true,"origin":"","legend":"\u003cp\u003eObserved (histograms) and fitted (curves) focal and benchmark distributions of composition indicators for reclaimed Brushland/Grassland, and their quality scores [lower, upper 95% CLs]. Pink: focal; blue: benchmark; purple: overlap.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/b419b5efee5e5bdbee8a90f6.png"},{"id":43843638,"identity":"3f94b6e8-49b8-4f64-bb6f-913b18fce418","added_by":"auto","created_at":"2023-09-28 15:50:19","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":92635,"visible":true,"origin":"","legend":"\u003cp\u003eObserved (histograms) and fitted (curves) focal and benchmark distributions and quality scores [lower, upper 95% CLs] of structural indicators for reclaimed Intermediate Forest, Early-mid. Pink: focal; blue: benchmark; purple: overlap.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/dc953cbeb59d540b139dc713.png"},{"id":43843639,"identity":"3723c819-6019-4430-9b13-709470a2b127","added_by":"auto","created_at":"2023-09-28 15:50:19","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":102453,"visible":true,"origin":"","legend":"\u003cp\u003eMonte Carlo simulations of power vs. sample size for overall quality for (a) four type of baseline vegetation and (b) their corresponding classes of reclaimed vegetation. Each curve is a logistic fit to a series of power-sample size simulations for a single vegetation type. Vertical dotted lines indicate the minimum sample sizes required to detect of change in overall quality of 0.1 at a=0.05 and a conventional power of 0.8.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/0d8e633291461c1a212c68ca.png"},{"id":53530885,"identity":"a07df09b-b3a1-469a-ba50-f20407fc7934","added_by":"auto","created_at":"2024-03-27 06:18:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":971723,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/222e5b7e-b81c-41ef-b4bc-0741bff63cbd.pdf"},{"id":43843667,"identity":"e550f2a1-5898-43b4-ae6d-6854a31f540e","added_by":"auto","created_at":"2023-09-28 15:50:20","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2192773,"visible":true,"origin":"","legend":"","description":"","filename":"BoyleetalSupplementaryMaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-3374961/v1/a414ea7d19eafc94a6508530.docx"}],"financialInterests":"Competing interest reported. BLB was funded in part by Teck during the preparation of this manuscript. Some of the paid consulting work performed by BLB and REG involves application of the methods described in the manuscript. AB and WF are employees of Teck.","formattedTitle":"Vegetation Quality Assessment: a sampling-based, loss-gain accounting framework for native, disturbed and reclaimed vegetation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIncreasing global awareness of the pervasive effects of human activity on the natural world (Ceballos et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Pimm and Joppa \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) is leading to voluntary and mandated approaches to reduce the impacts on biodiversity from resource development and related economic activities (International Finance Corporation \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rainey et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; CBD \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The extent and condition or quality of native vegetation is the principal currency by which impacts to terrestrial ecosystems are measured (Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). In addition to their inherent value as major components of the world's biodiversity (Kier et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), plants and plant communities provide the habitats and ecosystems upon which other organisms depend, and deliver essential environmental services through their influence on soil fertility, hydrological cycles and the global carbon balance (Schlesinger and Andrews \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). In the context of environmental impact assessment, inaccurate estimates of vegetation extent and quality can lead to reduced restoration effectiveness at impacted sites, mismatched biodiversity offsets, and potentially costly compliance failures (Maron et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). An accurate, objective and cost-effective framework for measuring vegetation quality is arguably one of the most important tools of regulators, developers, reclamation practitioners, and other stakeholders involved in monitoring and managing the effects of industrial development on terrestrial biodiversity (Parkes and Newell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2003\u003c/span\u003e, Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn many assessment and monitoring frameworks, vegetation \"quality\" refers to the degree to which vegetation at a site resembles native vegetation in the absence of human disturbance (Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Vegetation structured by natural disturbances (such as flooding and avalanches) or maintained by pre-European human intervention (such as fire-managed grasslands in North American and Australia) are also generally regarded as native vegetation (Parkes and Newell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Lewis et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Although the value of native vegetation \u003cem\u003eper se\u003c/em\u003e is uncontroversial (Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Landres et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), debate still surrounds the questions of how to measure the condition of vegetation which has been disturbed by human activity (McElhinny et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Cook et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) and how to select the appropriate reference vegetation to which the disturbed vegetation is compared and restored (Demeo et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs part of a broader program of biodiversity monitoring and management at Teck Coal Limited’s (Teck) Elk Valley steelmaking coal mining operations in southeastern British Columbia, Canada, we developed an accounting framework for quantifying impacts and improvements to native vegetation in terms of quality (as defined above) and quality hectares (quality x area; Rio Tinto \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; see Sahley et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2017\u003c/span\u003e for an example application). The framework focuses on natural ecosystems, with quality measured empirically relative to undisturbed native vegetation, and reflecting not only average conditions but also variation at the landscape scale. In terms of sensitivity, the method needed to be capable of distinguishing major vegetation types and successional (seral) stages, as well as different types of disturbance—all at temporal and spatial scales relevant to large mining operations. Field methods needed to be efficient, cost-effective, and—ideally—compatible with standard protocols to allow the re-use of existing vegetation data. Most importantly, the framework needed to be repeatable, verifiable, capable of estimating uncertainty, and minimally vulnerable to observer bias. Although initially developed in the specific context of Teck's Elk Valley mining operations, the framework and associated methodology is intended to be widely applicable, and is currently in use at a variety of projects involving multiple companies in Canada, USA, Suriname, Chile and Greece, and in vegetation ranging from boreal forest to desert scrub to lowland tropical rainforest. Here we describe this monitoring framework, which we call Vegetation Quality Assessment (VQA).\u003c/p\u003e \n\n, \n\n \n\n \n\n \n\u003ch3\u003eExisting loss-gain frameworks\u003c/h3\u003e\n\u003cp\u003eWe identified several existing approaches for quantifying losses and gains in vegetation or habitat quality. These included Habitat Hectares (Parkes and Newell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), BioCondition (Eyre et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), Biometric (Gibbons et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) and Quality Hectares (Rio Tinto \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), among others. These approaches typically measure vegetation quality by scoring ecological indicators relative to “benchmarks” representing the expected conditions in undisturbed native vegetation. Benchmark values are determined by sampling undisturbed vegetation adjacent to the project site or by reference to standard values for each vegetation or habitat type, as provided by regulatory agencies (e.g., The State of Queensland \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Multiple indicator scores are combined in various ways to produce an index of overall vegetation quality (Gibbons and Freudenberger \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Quality is then multiplied by area to provide a discounted measure of the amount and condition of vegetation to be compensated through restoration at the project site and offsets elsewhere. We follow Rio Tinto (Rio Tinto \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) in referring to quality-discounted area as “quality hectares” (QH).\u003c/p\u003e\u003cp\u003eAll existing methods failed to meet one or more of our key requirements. The most pervasive shortcoming was lack of a statistical framework. Sample plots were often located deliberately rather than at random, and replication, when mentioned at all, was on the order of 1–3 plots per vegetation type, with no explicit justification for the sample sizes used (e.g., Eyre et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In the absence of replication and randomization at the appropriate temporal and spatial scales, it is impossible to assess the degree to which the samples represent the site as a whole (Hurlbert \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1984\u003c/span\u003e). Furthermore, inadequate replication and lack of randomization increase vulnerability to observer bias; for example, field crews could accidentally or deliberately inflate quality scores by situating plots in unusually pristine sites. Another consequence of lack of replication is the inability to measure error or assess statistical significance. Quality is frequently reported without measures of variance (Parkes and Newell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Eyre et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) as are benchmark values provided by regulatory agencies (e.g., The State of Queensland \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA second shortcoming was the use of custom field protocols specific to each framework, thereby precluding the use of existing vegetation data. In many countries publicly-accessible repositories of vegetation data can be extensive, representing essentially all types of native vegetation (e.g., Peet et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). In the case of Teck's Elk Valley mining operations, the ability to leverage as reference data an existing provincial database of thousands of standardized vegetation plots (British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) was a key goal of our assessment framework. Without this critical resource we would have faced years of expensive data collection—a commitment of time and resources that many companies would find prohibitive. In addition, the sampling methodology commonly employed for vegetation plots within the provincial vegetation database (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) was also used by Teck over many years for mapping vegetation within the Elk Valley. The ability to re-purpose such pre-existing data for vegetation assessment would represent a tremendous saving of time and expense.\u003c/p\u003e\u003cp\u003eMany of the methods we reviewed omitted important indicators of vegetation condition. For example, BioCondition omit species identifications in favor of “key attributes or surrogates of biodiversity values that can be rapidly measured in the field” (Eyre et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), with the goal of enabling surveys to be \"undertaken rapidly by a range of natural resource managers…not just botanical ecologists\" (Parkes and Newell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). However, any potential gains in speed and ease enabled by such simplifications must be weighed against the information lost by ignoring taxonomic composition. Species composition is a fundamental attribute by which vegetation units are defined under most classification systems (e.g., Jennings et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), and indicators based on taxonomic identification are sensitive barometers of forest health, succession and disturbance. For example, small-scale disturbances can leave imprints on species richness that remain detectable for 100 years or more (Curtin \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Species composition discriminates at fine temporal scales among regenerating forests of different ages (Kappelle \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Major classes of temperate and tropical vegetation are readily distinguished by taxonomic composition, even at coarse taxonomic levels such as genus and family (Gentry \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Furthermore, the gains in efficiency enabled by discarding taxonomic identifications were not readily apparent. For example, the FS1333 Site Visit protocol used extensively in British Columbia (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) and employed for the vegetation plots included in the current study (see example application, below), includes identification and estimation of relative abundance of all plant species in the plot, in addition to structural and soil measurements. Yet the 1–2 hours taken by experienced practitioners to complete a “Site Visit” vegetation survey (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) is comparable to the 2 hours typically required for a BioCondition plot (Eddy et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), which omits most species-level identifications.\u003c/p\u003e\u003cp\u003eIn developing a vegetation quality framework within the specific context of the Elk Valley, we sought a method that would allow us to use data collected previously as part of Teck’s ongoing vegetation mapping and data collection program, and to leverage as benchmark data the vegetation inventories maintained by the British Columbia Ministry of Environment and Climate Change Strategy (BC ENV) and used to develop the provincial Biogeoclimatic Ecosystem Classification (British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). In addition, as our ultimate goal is to deploy the framework worldwide, we needed a method that would be applicable to all vegetation types, temperate and tropical, and compatible with a wide variety of commonly-used vegetation inventory methods.\u003c/p\u003e\n\u003ch3\u003eOverview of VQA\u003c/h3\u003e\n\u003cp\u003eThe framework we developed combines useful features of existing methods—including quality measured as similarity to undisturbed native vegetation and an accounting workflow based on QH—with a stratified-random sampling design that enables estimation of error and inference at the scope of the entire project. Losses due to project impacts and gains from restoration and offsets are measured relative to a baseline of vegetation quality and area prior to project initiation\u003ca class=\"FNLink\" href=\"#Fn1\" id=\"#FNLinkFn1\"\u003e\u003c/a\u003e. Multiple later assessments at project and offset sites track net quality hectares over time, with the ultimate goal of detecting the point at which baseline QH have been exceeded and NPI has been achieved (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). A key feature of the method is the measurement of quality in terms of overlap between indicator distributions of project site vegetation and in benchmark-quality vegetation of the same type. The approach can be applied to both forested or non-forested vegetation using widely-used vegetation sampling methods. Below, we outline the steps of the VQA workflow, followed by a trial implementation at Teck’s Elk Valley operations.\u003c/p\u003e\u003ch3\u003eLoss-gain accounting\u003c/h3\u003e\u003cp\u003eThe complete VQA workflow (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) consists of a series of loss-gain assessments of quality hectares of each vegetation type present at the project or offset site. Quality hectares is calculated as\u003c/p\u003e\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$QH=Q \\text{x} A$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cem\u003eQH\u003c/em\u003e is quality hectares, \u003cem\u003eQ\u003c/em\u003e is vegetation quality and \u003cem\u003eA\u003c/em\u003e is the area of that vegetation in hectares. As \u003cem\u003eQ\u003c/em\u003e varies from 0 to 1, \u003cem\u003eQH\u003c/em\u003e ranges from 0 to \u003cem\u003eA\u003c/em\u003e hectares.\u003c/p\u003e\u003cp\u003eThe workflow begins with a baseline assessment of QH for each class of vegetation prior to project start. At each subsequent assessment, the net change in QH relative to baseline (\u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet\u003c/em\u003e\u003c/sub\u003e) is determined by subtracting baseline quality hectares (\u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e) from current quality hectares (\u003cem\u003eQH\u003c/em\u003e) for the same vegetation:\u003c/p\u003e\u003cp\u003e\u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet\u003c/em\u003e\u003c/sub\u003e = \u003cem\u003eQH\u003c/em\u003e – \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e (2)\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTypically, with resource developments such as mines, not all losses at the project site can be restored. These residual impacts will need to be compensated by additional QH gained through conservation or restoration actions at one or more offset sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). Each offset site will require its own set of assessments, including a baseline assessment and separate calculations of net QH. Unlike the project site, offset assessments generally need to be adjusted for one or more counterfactual scenarios, such that only \u003cem\u003eadditional\u003c/em\u003e QH directly attributable to the offset program are included. In other words, existing QH that would have persisted into the future, or increases in QH that would have occurred in the absence of the offset program, must be excluded. The simplest counterfactual is complete elimination of an existing natural area. For example, if purchase and protection of an offset prevents a credible scenario of permanent conversion to agriculture resulting in complete loss of all native vegetation, then the baseline quality for the offset would be zero and net offset QH at each assessment would be simply the observed QH.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cp\u003eFor projects that include offsets, overall net QH is the sum of net QH at the project site and at each offset site. For a project with two offsets, net quality hectares is\u003c/p\u003e\u003cp\u003e\u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet\u003c/em\u003e\u003c/sub\u003e = \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.p\u003c/em\u003e\u003c/sub\u003e + \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.o1\u003c/em\u003e\u003c/sub\u003e + \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.o2\u003c/em\u003e\u003c/sub\u003e, (3)\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet\u003c/em\u003e\u003c/sub\u003e is overall net QH, \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.p\u003c/em\u003e\u003c/sub\u003e is net QH of the project site, and \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.o1\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet.o2\u003c/em\u003e\u003c/sub\u003e are net QH of offset sites 1 and 2, respectively. Each assessment of overall net QH thus includes losses due to project impacts, gains due to restoration and natural succession at the project site, and gains from averted losses and restoration at offset sites.\u003c/p\u003e\u003cp\u003eIn theory, attaining a NPI requires that \u003cem\u003eQH\u003c/em\u003e\u003csub\u003e\u003cem\u003enet\u003c/em\u003e\u003c/sub\u003e \u0026gt; 0 for all native vegetation types by the final assessment. However, each of the many measurements involved in measuring vegetation quality (see \"Confidence limits and significance testing\") has its own associated measurement error that propagates to all derived values. In VQA, each estimate of Quality, QH and QH\u003csub\u003enet\u003c/sub\u003e is a expressed as a mean with an associated 95% CI (see \"Confidence limits and significance testing\"). In practice, therefore, NPI cannot not be assumed until the lower 95% confidence limit of QH\u003csub\u003enet\u003c/sub\u003e exceeds 0 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003ch3\u003eStratification of vegetation, disturbance and reclamation\u003c/h3\u003e\u003cp\u003eEach assessment cycle requires a comprehensive classification and mapping of all land cover at the project site. Vegetation may be further subdivided into naturally occurring seral stages, if successional vegetation is a major component of natural landscapes in the region. For example, early successional forest can account for large areas in regions subject to frequent, naturally occurring fire (Fulé et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Korb et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Each class of vegetation (plus seral stage, if applicable) may be further subdivided into disturbance classes expected to follow divergent successional trajectories. Different restoration treatments should also be monitored separately if recovery outcomes and timelines are expected to differ, or if a goal of the program is to compare differences among treatments.\u003c/p\u003e\u003cp\u003eEach class of vegetation plus disturbance (and reclamation treatment, if applicable) represents a different sampling stratum. Within each stratum, the required number of sampling points (vegetation plots; see \"Effect size, power, and minimum sample size\") should be located at random using GIS software. If the initial classification results in a large number of strata, adequate sample sizes may be difficult or impossible to achieve given available time and resources. In such cases, it may be necessary to reduce the number of strata (thereby increasing per-stratum sample sizes) by combining ecologically-similar vegetation types.\u003c/p\u003e\u003cp\u003eEach assessment thus requires its own map of the project site, classified into strata of vegetation, disturbance and reclamation. Once the mapping process is complete, the actual area of each stratum is determined and multiplied by the quality of that stratum (see \"Quality calculations\", below) to obtain the stratum-specific current QH for that assessment. The QH of all strata sharing the same benchmark vegetation are summed to provide the total current QH for that class of native vegetation. In general, VQA seeks to reestablish vegetation of the same type that existed prior to project impacts; QH, net QH and NPI are therefore assessed independently for each class of benchmark vegetation (\"like-for-like\"; for a discussion of exceptions, see Text S1).\u003c/p\u003e\u003cp\u003eIdeally, the baseline map and baseline assessment are completed prior to initiation of project-related impacts. However, if the VQA framework is implemented after the project has begun, modeling (“back-casting”) can be used to reconstruct vegetation present prior to project impacts. This approach is described in detail in Knopff and Franklin (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003ea) and demonstrated in the Teck VQA application outlined below.\u003c/p\u003e\u003ch3\u003eVegetation sampling\u003c/h3\u003e\u003cp\u003eOnce mapping is complete and sampling strata have been defined, sampling localities for plots within strata are assigned randomly using GIS software. The number of plots per stratum can be estimated using Monte Carlo methods, as described below. We refer to vegetation plots from the project site as “focal plots”.\u003c/p\u003e\u003cp\u003eIn addition to focal plots, a reference sample of plots in high-quality, undisturbed examples of each type of native vegetation present at the project site is also required. These reference plots, or “benchmark plots”, can be established in areas adjacent to the project site (if undisturbed native vegetation is present), or obtained from existing vegetation inventory databases. Regardless of source, it is important that benchmark plots meet the following requirements: (a) indicators and sampling methods are identical or compatible with those of the focal plots, (b) the vegetation sampled can be reliably matched to native vegetation types present at the project site, and (c) sampling locations are minimally affected by human disturbance, based on accompanying metadata or expert evaluation. Ideally, experts not involved with the project should be responsible for verifying that plots or localities sampled represent undisturbed native vegetation. In calculating quality, different samples of focal plots representing distinct classes of disturbance or reclamation but the same native vegetation type can be compared to the same set of benchmark plots.\u003c/p\u003e\u003cp\u003eThe VQA framework is compatible with a variety of vegetation sampling methods. In many cases it may be possible to combine different field methods, as long as indicators measured are compatible. Legacy data collected using different protocols may be used if the relevant indicators can be transformed or otherwise standardized to produce equivalent measurements.\u003c/p\u003e\u003cp\u003eA final consideration influencing the choice of plot methods is the indicators to be measured. In most cases, a small number of vegetation attributes—in particular, species richness, taxonomic composition, growth form, and size or age class composition—are sufficient to distinguish among classes of succession and disturbance (Gentry \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Kappelle \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Chazdon et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Most commonly used vegetation sampling methods include such measurements and should be suitable for VQA.\u003c/p\u003e\u003ch3\u003eQuality calculations\u003c/h3\u003e\u003cp\u003eOverlap-based quality\u003c/p\u003e\u003cp\u003eThe principal measure of quality within the VQA framework is overlap between the probability distributions of indicators in the focal and benchmark vegetation. Measuring similarity in terms of overlap between probability distributions has a long history in economics (e.g., comparison of incomes; Gastwirth, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1975\u003c/span\u003e). The overlap method accommodates a wide variety of normal and non-normal distributions, and deserves to be more widely used—especially in ecology, where skewed and bounded distributions are the norm rather than the exception (Magurran \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Overlap between focal and benchmark indicator distributions provides an intuitively meaningful index of similarity ranging from 0 to 1, where 1 (100% overlap) means, effectively, \"indistinguishable from benchmark\". Furthermore, distribution overlap captures the full range of natural variation, including asymmetry not evident in comparisons of means (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo calculate overlap, each indicator must first be assigned a distribution family based on the domain of the variable and visual inspection. Appropriateness of the distribution can be formally assessed using a goodness-of-fit test such as the one-sample Kolmogorov-Smirnov test (Berger and Zhou \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Commonly measured indicators generally fall into one of three distributions: (1) negative binomial (counts such as species richness), (2) gamma (zero-bounded continuous measurements such as taxonomic similarity/dissimilarity), and (3) beta (percentages or proportions such as percent cover of species or growth forms). As the negative-binomial is a discrete distribution, it is approximated by the gamma distribution when calculating distribution overlap (Best and Gipps \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e1974\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOverlap is calculated in a two-step process. In step one, separate univariate pdfs are fit to the focal and benchmark samples using maximum likelihood estimation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea; see Venables and Ripley \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). In step two, the area of the intersection of the two distributions is determined by integration. Confidence limits (CLs) of the overlap-quality estimate are determined by bootstrapping these calculations, using the bootstrap deviance approach described below (see \"Confidence limits and hypothesis testing\").\u003c/p\u003e\u003cp\u003eAlternative quality methods\u003c/p\u003e\u003cp\u003eAlthough the \"overlap quality\" method described above accommodates the majority of indicators, three alternative approaches are available to handle special cases. The first of these, one-tailed quality, calculates quality using values either above or below the benchmark mean, but not both. Depending on the goals of the application, one-tailed quality may be appropriate for indicators such as Species Richness if high values of richness are considered desirable; in this case, quality would be calculated using values below the benchmark mean (i.e., the \"lower tail\"). Details of the one-tailed quality algorithm are provided in Text S2 and Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cp\u003eFixed-benchmark quality, which applies to beta-distributed indicators only, uses a single fixed benchmark value. Quality is calculated as the absolute value of the difference between the benchmark value and the focal mean, divided by the maximum possible difference on the [0,1] domain, with an additional, non-linear transformation that causes quality to drop steeply with increasing distance from benchmark (Fig. S2). Fixed-benchmark quality is appropriate when a single, ideal benchmark value is readily apparent and appropriate. Examples of indicators for which a fixed benchmark might be used include percent cover of invasive species and infection rates of diseases (e.g., \"Percent of trees infected White Pine Blister Rust\"); in both cases, the targeted benchmark is 0. Details of the calculation of fixed-benchmark quality are provided in Text S3 and Fig. S2.\u003c/p\u003e\u003cp\u003eThe third alternative, overlap-means quality, combines features of the overlap and fixed-benchmark algorithms to resolve problems caused by zero and one inflation in beta distributed indicators. Zero- and one-inflated distributions are common among ecological indicators measured as percentages (Magurran \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Such mixed distributions can distort measures of quality based purely on distribution fitting. The overlap-means algorithm calculates quality twice, once based on distribution overlap and a second time based on the difference between the focal and benchmark mean, and uses a weighted average of the two. The result is a robust quality score which provides an intuitively accurate representation of the similarity of two distributions across the entire beta domain, even for mixed distributions. For details of the calculation of overlap-means quality, see Text S4 and Figs. S3-S4.\u003c/p\u003e\u003cp\u003eOverall quality\u003c/p\u003e\u003cp\u003eMost existing vegetation quality methods combine indicator scores into a single index of overall quality. To reduce the influence of correlated indicators and minimize subjective weightings (see critique in McCarthy and Parris \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), we group related indicators into “functional groups” representing key aspects of ecological function. For example, indicators based on taxonomic composition, such as species richness or taxonomic composition, are assigned to functional group “Composition”, whereas indicators that measure the relative abundance of different growth forms, such as “percent Cover Herbs” and “percent Cover Trees”, are assigned to functional group “Structure”. For each functional group, functional group quality is calculated as the arithmetic mean of all indicator qualities within that functional group. Overall quality is then calculated as the geometric mean of the functional group qualities, as follows:\u003c/p\u003e\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$Q= {\\left(\\prod _{i=1}^{N}Q{f}_{i}\\right)}^{\\frac{1}{N}},$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e8\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cem\u003eQ\u003c/em\u003e is overall quality, \u003cem\u003eQf\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e is the arithmetic mean of indicator qualities in functional group \u003cem\u003ei\u003c/em\u003e, and \u003cem\u003eN\u003c/em\u003e is the number of functional groups.\u003c/p\u003e\u003cp\u003eThe use of the geometric mean addresses the criticism that existing approaches are vulnerable to manipulation of quality scores through the substitution of one indicator for another (McCarthy and Parris \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). For example, in quality frameworks that score indicator “downed dead wood” as a habitat feature, it is possible to inflate quality while harvesting all trees if some of trees are left on the ground as \"downed dead wood\" (McCarthy and Parris \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Under the VQA approach, a single functional group whose mean drops to zero will force overall quality to zero as well, regardless of the values of the other functional group means. This behavior ensures that all key components of ecological function have non-zero quality. At the same time, use of the arithmetic mean for functional group quality prevents indicators within the same group from forcing overall quality to zero during early succession, when cover (and quality) of herbs and shrubs recovers long before cover (and quality) of large trees becomes non-zero.\u003c/p\u003e\u003cp\u003eConfidence limits and significance testing\u003c/p\u003e\u003cp\u003eIn VQA, the 95% CLs of indicator quality, functional group quality, overall quality and QH are estimated using the bootstrap deviance method of Efron (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e1979\u003c/span\u003e). Each quality calculation is repeated a large number of times (generally, 10,000 iterations), sampling the actual data with replacement. The lower 95% CL is calculated by subtracting the 0.025 quantile of the bootstrap deviances (i.e., deviance of the bootstrap estimates from the bootstrap mean) from the observed overlap. Likewise, the upper 95% CL is calculated by adding the 0.975 quantile of the bootstrap deviances to the observed overlap. This method accurately represents asymmetric confidence intervals (CIs) that commonly result from overlap of distributions such as the beta distribution.\u003c/p\u003e\u003cp\u003eUnder certain circumstances, 95% CLs can be used directly to evaluate significant differences. For example, the difference between a measured value of quality and a fixed target value is significant at p ≤ 0.05 if the fixed value lies outside the 95% CLs of the observed distribution. The difference between two estimates of quality or quality hectares is significant if the CIs of the difference of the estimates do not overlap 0 (Cousineau \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In VQA, we assess progress to NPI by examining the position of the lower 95% CL of the difference between current and baseline QH (i.e., Net QH). Once the lower CL exceeds 0, we can be 95% confident that QH\u003csub\u003ecurrent\u003c/sub\u003e \u0026gt; QH\u003csub\u003ebaseline\u003c/sub\u003e and NPI has been attained, given the variability of our estimates of Quality and QH (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb).\u003c/p\u003e\u003cp\u003eEffect size, power, and minimum sample size\u003c/p\u003e\u003cp\u003eThe difference between the mean and the 95% CL, or margin of error (MOE), is an intuitive measure of effect size, precision and error which deserves to be more widely used (Cousineau \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Because MOEs calculated using the bootstrap deviance method are frequently asymmetrical, we use the largest of the two MOEs of quality as a conservative measure of the \"minimum detectable effect size\" (MDES) attainable at a given sample size. Defining effect size in this way, we can use Monte Carlo methods to estimate the minimum sample size (\u003cem\u003en.min\u003c/em\u003e) required to detect a given target MDES (\u003cem\u003eMDES\u003c/em\u003e\u003csub\u003e\u003cem\u003eexp\u003c/em\u003e\u003c/sub\u003e) at \u003cem\u003eα\u003c/em\u003e = 0.05 and a conventional power of 0.80 (see Text S5 for details). This approach is illustrated in the example application below. Throughout, we follow the recommendations of Cumming (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) and Cousineau (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) in presenting upper and lower CLs in square brackets.\u003c/p\u003e\n\u003ch3\u003eExample application: A VQA accounting of baseline \u0026 reclaimed vegetation at Teck Elk Valley Mine Operations\u003c/h3\u003e\n\u003ch2\u003eGoals of the application\u003c/h2\u003e\u003cp\u003eAs a trial application of the VQA framework, we performed assessments of baseline and current reclaimed vegetation quality for Teck’s steelmaking coal mining operations in the Elk Valley of British Columbia, Canada. The baseline assessment estimated quality and extent of all vegetation at the start of five Elk Valley operations, prior to mining-related impacts. The current assessment measured present-day extent and condition of reclaimed vegetation within the study area. The main goals of the application were four. The first goal was to demonstrate the retroactive application of the VQA framework, using modeling approaches to reconstruct terrain and vegetation originally present within an area already impacted by mining activities (see Knopff and Franklin \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e for details). The second goal was to develop a best estimate of the baseline extent and condition of all native vegetation impacted by Teck’s operations in the Elk Valley over the period 1950–2018. The third goal was to demonstrate the use of existing vegetation data within the VQA framework by combining focal data from vegetation surveys previously collected from the project site with benchmark data from a public vegetation inventory database. The fourth goal was to estimate the current contribution of reclamation to NPI, with the aspirational goal of ensuring that gains from reclamation, revegetation and biodiversity offsets ultimately exceed impacts to native vegetation within the combined mine footprints.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eStudy site\u003c/p\u003e\u003cp\u003eThe study area encompassed the combined past and current footprints of all Teck steelmaking coal mining operations within the Elk River Valley (“Elk Valley”) of southeast British Columbia. The Elk Valley is located along the western edge of the Rocky Mountains, in terrain dominated by steep slopes and narrow valley bottoms. The valleys are generally underlain by sandy and shale rocks of the Jurassic and Cretaceous eras. Well-drained soils usually occur on steep slopes with a thin layer of surficial material over bedrock. Areas with poor drainage occur on lower and toe slope positions and in isolated areas along nearly level terrain. The Elk Valley is largely forested, mainly of coniferous stands. The most common Biogeoclimatic zones include the Engelmann Spruce–Subalpine Fir Dry Cool variant and Engelmann Spruce–Subalpine Fir Dry Cool Woodland subzone (Meidinger and Pojar \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1991\u003c/span\u003e). Teck’s steelmaking coal operations within the Elk River Valley include four active mines and one under care and maintenance and have a combined area of c. 19,000 ha. This total includes areas currently disturbed by mining plus areas that were disturbed in the past but are now reclaimed or undergoing reclamation, through 2018.\u003c/p\u003e\u003cp\u003eVegetation mapping and classification\u003c/p\u003e\u003cp\u003eFor the baseline assessment, vegetation of the mine footprint was reconstructed using a combination of archival aerial photos, satellite imagery, and Predictive Ecosystem Mapping (PEM; Terrestrial Ecosystem Mapping Alternatives Task Force Resource Inventory Committee (RIC) \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e1999\u003c/span\u003e) as well as terrain reconstruction and hydrological and soil models. The resulting back-cast vegetation polygons were combined with maps of existing vegetation to produce a comprehensive baseline map of pre-mine vegetation conditions within all Elk Valley operations administered by Teck (Knopff and Franklin \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). All vegetation units were classified to the Site Series level, using the Biogeoclimatic Ecosystem Classification (BEC) version 10 (British Columbia Ministry of Forests-Lands and Natural Resource Operations \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Seral stage classes were assigned to forested vegetation using BC Ministry of Environment and Climate Change Strategy standard terminology (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAs detailed information on vegetation condition at the baseline year was not available, we used the condition of current vegetation outside the mine footprint to represent vegetation condition within the mine footprints at baseline. We believe this assumption is conservative as land management practices have generally improved over time and anthropogenic fire along with other large-scale land disturbance practices have decreased; therefore, current condition should be as good as, or better than, condition at the baseline year (although we acknowledge that fire may have contributed to increased quality for drier forest types which were structured by fire in pre-European contact times). For much of the 20th century, the extent of old growth forest within the Elk Valley was considerably less than under historical disturbance regimes prior to European settlement in the mid-1850s. In particular, the lack of old growth Montane Spruce (MS) and Interior Cedar Hemlock (ICH) forest reflects a combination of very large and severe fires in 1919 and in the 1930s, forest clearing for urban/rural development and agriculture, and timber harvesting that has been ongoing since the early 1900s (Holmes et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). As we did not subdivide native vegetation into subcategories of disturbance, each vegetation type is represented by exactly one focal land cover class only.\u003c/p\u003e\u003cp\u003eThe resulting vegetation and land cover map for the study site consisted of 133 site series (MacKenzie \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; British Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Including structural stages of forested vegetation, the final number of distinct land cover classes (sites series + structural stages) was 396. As this number of sampling strata was too large to be sampled at adequate sample sizes, we revised the baseline map, combining site series into a smaller number of classes (hereafter referred to as \"ecosystem groups\") representing vegetation of similar structure, composition and moisture availability. For example, site series representing above-treeline alpine vegetation were combined into the ecosystem group “Alpine”. In addition, for forested vegetation, the 14 BEC structural stages were combined into the three seral stages “Early-mid”, “Mature” and “Old”. After these simplifications, the final vegetation and land cover map consisted of 21 ecosystem groups plus seral stage sampling strata (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Five sampling strata (Alpine grassland, Alpine meadow, Deciduous floodplain, Herb meadow and Shrubland) had focal or benchmark sample sizes that were too low (0–4 plots) to allow analysis. The total area of all units within the footprint study area was 18,978.9 ha.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBenchmark and focal plot sample sizes for all vegetation classes (ecosystem groups) within the study area, for the baseline and current (reclamation) assessments. Sample sizes will be smaller than shown for individual indicators not measured in all plots.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eFocal plots\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEcosystem group\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBenchmark plots\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBaseline native vegetation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCurrent reclaimed vegetation\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine Dwarf Shrub\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine grassland\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine meadow\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvalanche feature\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBrushland/Grassland\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e98\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeciduous floodplain\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, Early-mid\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e79\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, Mature\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e55\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, Old\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHerb meadow\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, Early-mid\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e148\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e220\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, Mature\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e83\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e216\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, Old\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKrummholz\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRock/Talus\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e64\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShrubland\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, Early-mid\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, Mature\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e77\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e54\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, Old\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWetland\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFor the current vegetation assessment, all reclaimed vegetation types within the footprint were assigned a target benchmark vegetation and the total area of each reclamation class was determined. Reclaimed vegetation fell into four ecosystem groups: “Brushland/Grassland”, “Early-mid Dry Forest”, “Early-mid Intermediate Forest” and “Early-mid Wet Forest”. As our accounting encompassed past, present and future impacts through 2035, including planned expansions, vegetated areas currently falling within the footprint of future mine expansions were coded as \"non-vegetated\" and included with the un-reclaimed areas within the current mine footprint. Total area of reclaimed vegetation was 2,109.0 ha, with the remaining non-vegetated areas totaling 16,869.9 ha.\u003c/p\u003e\u003cp\u003eVegetation plots\u003c/p\u003e\u003cp\u003eThe baseline vegetation plots were sampled over a 41-year period post-dating the baseline year, from 1975–2016, in native vegetation adjacent to the project site (i.e., the current mine footprint). These samples provide a conservative estimate of the condition of native vegetation present prior to mining impacts (see \"Vegetation mapping and classification\"). These data were collected by multiple consulting companies as part of Teck’s ongoing vegetation mapping and inventory program within the Elk Valley. After removing plots not compatible with BC ENV Site Visit methodology (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) the final baseline focal dataset consisted of 1,121 vegetation plots.\u003c/p\u003e\u003cp\u003eVegetation plots for the current assessment were sampled in reclaimed vegetation within the project footprint. These plots are located in multiple areas representing different reclamation treatments. For many of them, reclamation was initiated more than a decade ago. Unlike current practices, many historical reclamation activities did not have restoration of native vegetation as a primary goal, and included planting of non-native agronomic species, particularly in grassland ecosystems targeted to support an ungulate end land use. Despite potential differences in reclamation methods and initiation times, we did not distinguish among different reclamation treatments. Our goal for this preliminary VQA application was to gain a general understanding of the current state of historical reclamation for broad classes of vegetation. This information will be used to plan more fine-grained sampling in future assessments, and enable adaptation of reclamation practices to better serve Teck's current focus on native vegetation. All current assessment focal plots were collected using the same Site Visit methodology as the baseline assessment plots. The final reclamation data set consisted of 222 plots.\u003c/p\u003e\u003cp\u003eBenchmark plots were obtained from a database of inventories developed by the BC ENV as reference data for the BEC vegetation classification standard (Meidinger and Pojar \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; MacKenzie \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; British Columbia Ministry of Forests-Lands and Natural Resource Operations \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). These data were also recorded using the BC ENV Site Visit methodology (BC Ministry of the Environment \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). After selecting the subset of inventories which matched at the Site Series level to native vegetation units present within the focal site baseline map (see \"Vegetation mapping and classification\"), the final benchmark dataset consisted of 959 vegetation plots. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the distribution of the benchmark, baseline and reclamation vegetation plots among the 21 ecosystem groups.\u003c/p\u003e\u003cp\u003eEcological Indicators\u003c/p\u003e\u003cp\u003eWe selected ten attributes that were recorded with sufficient consistency across all data sets to be used as ecological indicators (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Three of these (“Species richness”, \"Taxonomic distance\" and “Percent cover exotic species”) were taxonomic-based indicators which we combined within functional group \"Composition\". Four (“Percent cover herbs”, “Percent cover moss”, “Percent cover shrubs”, “Percent cover trees”) reflected aspects of vegetation height and complexity and were therefore placed in functional group \"Structure\". The remaining three (“Percent cover dead wood”, “Percent cover organic soil” and “Percent cover surface water”) were aspects of the physical environment which influence vegetation composition and structure; these were assigned to functional group \"Environment\". Indicator “Percent cover surface water” was included only in quality calculations for wetland vegetation (ecosystem group “Wetland”). “Species richness” was the count of all native species recorded in the plot. Native and exotic (introduced) species status was assigned using the 2016 edition of the BC Flora Checklist (BC Ministry of the Environment \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Indicator “Taxonomic distance” (TD) measured dissimilarity of species composition of the focal plots relative to the benchmark plots, as determined using non-metric multidimensional scaling (NMDS); details of the calculation of this indicator are provided in Text S6 and Fig. S5. The remaining indicators were measured as described in BC Ministry of the Environment (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eEcological indicators used for the example VQA application, their probability distributions and quality scoring methods.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndicator\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDistribution\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eQuality algorithm\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTest tail\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFunctional group\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpecies Richness\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNegative binomial\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLower\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eComposition\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTaxonomic Distance\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGamma\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eComposition\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Exotic Species\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFixed (0)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003en/a\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eComposition\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Herbs\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStructure\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Moss\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStructure\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Shrubs\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStructure\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Trees\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStructure\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Decomposing Wood\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEnvironment\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Organic Soil\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEnvironment\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercent Cover Surface Water*\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBeta\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverlap-means\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBoth\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEnvironment\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e*Indicator “Percent cover surface water” used for “Wetland” vegetation only.\u003c/p\u003e\u003cp\u003eQuality and quality hectares\u003c/p\u003e\u003cp\u003eWe calculated indicator quality and overall quality for sampling strata with at least 6 focal and 6 benchmark plots. Quality of “Percent cover exotic species” was calculated using a fixed benchmark value of 0. All other indicators used the overlap or overlap-means quality algorithms (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The 95% CLs of quality were calculated using 10,000 bootstrap iterations. QH and QH.net were calculated following the procedure described in \"Loss-gain accounting\". Vegetation types without at least 6 focal and 6 benchmark plots were excluded from quality calculations. For these vegetation types, we conservatively assumed a baseline quality of 100% and therefore used their actual areas as Quality Hectares in all subsequent calculations.\u003c/p\u003e\u003cp\u003eEffect size, power, and sample size\u003c/p\u003e\u003cp\u003eWe performed Monte Carlo simulations for each vegetation type of the relationship between sample size (\u003cem\u003en\u003c/em\u003e; equal for focal and benchmark) and power to detect a significant change in overall quality at \u003cem\u003eα\u003c/em\u003e = 0.05 and an effect size (MDES) of 0.1, as described in \"Effect size, power, and minimum sample size\". For each class of vegetation, the required minimum sample size, \u003cem\u003en\u003c/em\u003e\u003csub\u003e\u003cem\u003emin\u003c/em\u003e\u003c/sub\u003e, was the value of \u003cem\u003en\u003c/em\u003e at which a logistic fit to a plot of \u003cem\u003en\u003c/em\u003e and power for all simulations crossed the conventional power line of 0.80. We also performed power-sample size simulations for selected indicator-vegetation combinations. As some indicators required very large minimum sample sizes to achieve an MDES of 0.1, we also ran simulations at MDESs of 0.15 and 0.2 to examine the trade-off between minimum sample size and sensitivity.\u003c/p\u003e\u003cp\u003eSoftware\u003c/p\u003e\u003cp\u003eAll VQA calculations were implemented in the R programming language (R Development Core Team 2008). Univariate pdfs were fit to benchmark and focal samples by maximum likelihood estimation using function fitdistr of package MASS (Ripley et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e"},{"header":"Results and discussion","content":"\u003cp\u003eBaseline assessment\u003c/p\u003e \u003cp\u003eBaseline overall quality [lower, upper 95% CLs] of the assessed vegetation types ranged from a low of 0.74 [0.64, 0.88] for Krummholz to a high of 0.87 [0.83\u0026ndash;0.95] for Intermediate forest, Mature (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e; SM-1). The below-benchmark overall quality scores of forested vegetation, plus details of individual indicators (see below) were consistent with a history of logging in the early 1900s, prior to the initiation of the mines. Distributions of individual indicators and their quality scores (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) support this interpretation. For early-mid seral stages of all three forest types, a disproportionate number of focal sites had little to no large tree cover and consequently low quality scores for indicator Percent Cover Trees (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea for Intermediate Forest, Early-mid; results for Early-mid Dry Forest and Wet Forest similar). These anomalies were structural only; taxonomic composition of early-mid seral forests was almost identical to benchmark (e.g., Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). In addition, strongly bimodal focal distributions of indicator Percent Cover Trees suggest that the Early-mid seral forest class in the Elk Valley may represent mixtures of naturally regenerating forest and recovering forest harvest. If true, the accuracy of the baseline assessment would be improved by splitting each stratum into two disturbance classes, one natural and one disturbed, sharing the same benchmark vegetation, and assessing each class independently with a separate set of focal plots. This conclusion highlights the value of examining individual indicator distributions and quality scores in addition to overall quality.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline and current (reclaimed) vegetation quality for the combined Teck Elk Valley Operations footprint. For vegetation for which quality could not be calculated due to insufficient sample size (indicated by *), 100% quality is assumed. Total footprint area (all ecosystem groups combined) is 18,978.9 ha. CLs: confidence limits.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEcosystem group\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBaseline quality [95% CLs]\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eReclaimed (current) quality [95% CLs]\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.79 [0.72\u0026ndash;0.88]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine dwarf shrub\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.76 [0.68\u0026ndash;0.85]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine grassland*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine meadow*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvalanche feature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.86 [0.81\u0026ndash;0.95]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBrushland/Grassland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.82 [0.78\u0026ndash;0.88]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.35 [0.32\u0026ndash;0.40]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeciduous floodplain*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.81 [0.76\u0026ndash;0.90]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.42 [0.39\u0026ndash;0.46]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.84 [0.79\u0026ndash;0.94]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, old*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHerb meadow*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.79 [0.76\u0026ndash;0.91]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.31 [0.28\u0026ndash;0.36]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.87 [0.83\u0026ndash;0.95]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, old\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.76 [0.69\u0026ndash;0.83]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKrummholz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.74 [0.64\u0026ndash;0.88]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRock/Talus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.82 [0.77\u0026ndash;0.89]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShrubland*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.75 [0.70\u0026ndash;0.83]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.26 [0.19\u0026ndash;0.32]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.81 [0.75\u0026ndash;0.90]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, old\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.82 [0.74\u0026ndash;0.90]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWetland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.83 [0.78\u0026ndash;0.90]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eNon-forested vegetation classes also had overall quality scores\u0026thinsp;\u0026lt;\u0026thinsp;100% (i.e., CIs not overlapping 100%); thus, factors other than forest harvesting were responsible for the below-benchmark baseline vegetation quality. In the case of non-forested alpine vegetation (Alpine, Alpine Dwarf Shrub, Avalanche Feature, Krummholz), the somewhat lower quality scores for structural and environmental indicators (Table S2) are consistent with the moderate structural degradation noted by previous studies and attributed to intense grazing by native ungulates (eg, Poole et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). At lower elevations, quality of indicator \"Percent Cover Exotic Species\u0026rdquo; for Brushland/Grassland (0.77 [0.66, 0.85]) was lower than other forested and non-forested vegetation type (all of which had CIs close to or overlapping 100%; see Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Although the majority of plots in Brushland/Grassland had zero cover of exotic species, a few had high abundance, with exotic cover at some focal sites exceeding 60% (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). A number of benchmark plots (not used to calculate this indicator; see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) also had exotic species cover as high as ~\u0026thinsp;30%. The prevalence of exotic species in Brushland/Grassland is also reflected in its somewhat anomalous species composition and low score for Taxonomic Distance (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). These observations are consistent with a general pattern of invasion of western North American grasslands and shrublands by exotic species (Dukes and Mooney \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eReclamation assessment\u003c/p\u003e \u003cp\u003eOf the original 18,978.9 ha within the study area mine footprint, 2,109.0 ha had been re-vegetated at time of this analysis to the four reclamation vegetation classes shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. Overall quality of reclaimed vegetation ranged from 0.26 [0.19\u0026ndash;0.32] for Wet forest, Early-mid to 0.42 [0.39\u0026ndash;0.46] for Dry Forest, Early-mid, with no upper CLs exceeding 0.46.\u003c/p\u003e \u003cp\u003eQuality of the individual indicators paints a clearer picture of the ways in which each type of historical reclamation differed from undisturbed vegetation. Brushland/Grassland had fewer species, different taxonomic composition and much higher cover of non-native species, relative to the benchmark condition (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e; see also Table S3). These results are consistent with the widespread use of a small number of non-native agronomic species in historical grassland reclamation, which was guided largely by provincially-approved standards for ungulate forage production. Reclaimed forest differed strongly from naturally occurring early-mid seral forest in both structure (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e) and composition (Tables S3 \u0026amp; S4), with anomalously low cover of mosses, shrubs and trees (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003eb-d), lower species richness and different taxonomic composition (Fig. S6).\u003c/p\u003e \u003cp\u003eSample size requirements\u003c/p\u003e \u003cp\u003eSimulations of power versus sample size indicated that sample sizes ranging from 9\u0026ndash;17 were required to detect a change in overall quality of 0.1 in the baseline vegetation (e.g., Fig.\u0026nbsp;12). By contrast, much smaller sample sizes were needed to detect a similar effect size in the reclaimed vegetation (Fig.\u0026nbsp;11a). This difference reflects the much narrower CIs of overall quality in the reclaimed vegetation (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), which in turn are a consequence of the low variance of many reclaimed vegetation indicators (i.e., Figs.\u0026nbsp;9b, 9c, 9d and 10a; note especially the predominance of values close to or exactly equal to zero). The low minimum sample sizes shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb should not be taken literally; considerations of power and effect size aside; sample sizes as small as n\u0026thinsp;=\u0026thinsp;3 are unlikely to be representative of an entire class of reclamation. Furthermore, as reclaimed vegetation matures and the variances of its indicators and overall quality converge upon values observed in the benchmark vegetation, \u003cem\u003en.min\u003c/em\u003e should also increase. Ultimately, minimum sample sizes for mature reclaimed vegetation will be similar to the values of \u003cem\u003en.min\u003c/em\u003e determined for the baseline vegetation (e.g., Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). In practice, smaller samples of \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6 plots can be installed initially for monitoring reclamation quality during the first few years, and additional plots added incrementally as the vegetation matures.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSample size requirements for individual indicators were generally higher than for overall quality. For example, samples sizes of 18, 27 and 42 were required to achieve effect sizes of 0.2, 0.15 and 0.1, respectively, for Species Richness of vegetation \"Avalanche Feature\", (Fig. S7). Given the importance of the indicator distributions and indicator quality for diagnosing and resolving causes of low overall quality scores, we recommend larger sample sizes than the minimum required for the measurement of overall quality. Finally, we must emphasize that simulation-based estimates of minimum samples sizes are approximations only, and should be evaluated in light of the actual CLs obtained and by repeating power simulations as more data become available. If an acceptable level of error is attained prior to reaching the estimated minimum sample size, no additional sampling is needed.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAn even more compelling reason for targeting higher minimum sample sizes than those suggested here is to minimize the amount by which the estimated current QH must exceed baseline QH to ensure that the lower 95% CL of Net QH is greater than zero (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The decision to target a more stringent MDES of 0.05 or lower by increasing sample sizes can be made on a case by cases basis by weighing the costs of increased sampling versus additional restoration and offset obligations.\u003c/p\u003e \u003cp\u003eQuality Hectares and progress to NPI\u003c/p\u003e \u003cp\u003eComparison of current to baseline QHs shows the contribution of historical reclamation to the overall goal of NPI (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Although impacts to most classes of vegetation will need to be compensated via offset actions elsewhere, on-site restoration has made major contributions toward NPI for Brushland/Grassland and Early-mid Dry Forest; 62% and 45%, respectively, of the baseline QH of these two vegetation types had been restored by the time these data were collected. In addition to extending re-vegetation efforts to areas not currently undergoing reclamation, additional gains in QH could be achieved by improving the quality of existing and future reclamation. This is particularly true of Brushland/Grassland, where replacement of non-native grasses with native species has the potential to provide substantial improvements in quality for multiple indicators (see \"Reclamation assessment\"). In the case of Early-mid Dry Forest, this vegetation is expected to transition over time to Mature Dry Forest, and restoration should therefore aim to restore substantially more than the number of QHs originally present of the early-mid seral stage. Over-planting of the earliest seral stages will ensure a more balanced restoration of seral stages at and beyond closure. Teck has recently refocused the current reclamation program to target native species as well as aiming towards natural ecosystem assemblages for end land use objectives (Franklin and Burton \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). In this context, VQA provides a mechanism for integrating the contribution of native species and natural ecosystem assemblages into the goal of NPI.\u003c/p\u003e \u003cp\u003eFinally, it is important to emphasize that CLs, not means, determine when NPI has been attained. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, only when the lower 95% CL of Net QH exceeds zero (QH.net.lcl\u0026thinsp;\u0026gt;\u0026thinsp;0) are we justified in concluding that our estimate of QH.net also exceeds zero. It should also be apparent that excessively broad CIs can increase restoration obligations beyond what would be required given a more accurate estimate of QH.net. The incorporation of measurement error into the VQA workflow thus creates a built-in incentive for estimating quality, QH, and net QH with the greatest possible accuracy, subject to the opposing constraint of increased sampling and monitoring costs.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline area and Quality Hectares (QH), current (reclaimed) area and QH, and net QH of vegetation within the combined Teck Elk Valley Operations footprint. Baseline and current quality hectares are the product of the areas shown here and the quality scores in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. Net QH is current QH - baseline QH. For vegetation for which quality could not be calculated due to low sample size (indicated by asterisk), baseline QH\u0026thinsp;=\u0026thinsp;actual area (ha). CLs: confidence limits.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026minus;\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVegetation (ecosystem group)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eBaseline\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eCurrent (reclaimed)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eNet QH\u003c/p\u003e \u003cp\u003e[95% CLs]\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea\u003c/p\u003e \u003cp\u003e(ha)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eQuality Hectares\u003c/p\u003e \u003cp\u003e[95% CLs]\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea\u003c/p\u003e \u003cp\u003e(ha)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eQuality Hectares\u003c/p\u003e \u003cp\u003e[95% CLs]\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e385.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e304.2 [277.2, 338.8]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-304.2 [-338.8, -277.2]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine dwarf shrub\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.5 [0.41, 0.51]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-0.5 [-0.5, -0.4]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine grassland*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e67.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e67.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-67.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlpine meadow*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvalanche feature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e741.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e637.7 [600.6, 704.4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-637.7 [-704.4, -600.6]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBrushland/Grassland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e862.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e706.8 [672.4, 758.6]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1257.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e440.2 [402.5, 503.1]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-266.6 [-269.9, -255.5]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeciduous floodplain*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e157.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e157.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-157.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e920.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e745.5 [699.5, 828.4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e800.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e336.3 [312.3, 368.4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-409.2 [-460, -387.2]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e521.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e438.0 [411.9, 490.1]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-438 [-490.1, -411.9]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry forest, old*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e94.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e94.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-94.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHerb meadow*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7,771.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6139.1 [5,906.0, 7,071.6]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.1 [1.0, 1.3]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-6,138 [-7,070.3, -5,905]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3,931.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3420 [3,262.7, 3,734.5]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-3,420 [-3,734.5, -3,262.7]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate forest, old\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,272.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e966.7 [877.7, 1,055.8]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-966.7 [-1,055.8, -877.7]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKrummholz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e55.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e41.1 [35.6, 48.9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-41.1 [-48.9, -35.6]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRock/Talus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e233.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e191.4 [179.7, 207.7]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-191.4 [-207.7, -179.7]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShrubland*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-9.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, early-mid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e923.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e692.3 [646.1, 766.1]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e47.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e12.2 [8.9, 15.0]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-680.1 [-751.1, -637.2]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, mature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e565.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e457.7 [423.8, 508.5]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-457.7 [-508.5, -423.8]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet forest, old\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e317.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e259.9 [234.6, 285.3]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-259.9 [-285.3, -234.6]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWetland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e138.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e115.2 [108.3, 124.9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e-115.2 [-124.9, -108.3]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe VQA approach provides a quantitative, objective framework for measuring the extent and quality of pre-impact vegetation, determining compensation obligations, monitoring the pace and effectiveness of restoration, and determining the point at which NPI has been achieved. By using distribution overlap as the main measure of quality, the method incorporates natural, landscape-scale variation not captured by other methods. Examination of the distributions of individual indicators provides insights that can be used to fine-tune restoration practices and adjust offset goals. Random sampling reduces vulnerability to bias and enables statistical inference at the scale of entire landscape. The ability to quantify uncertainty increases confidence in monitoring results and creates built-in incentives for maximizing accuracy of quality estimates.\u003c/p\u003e \u003cp\u003eA potential shortcoming of the method is the need for relatively large sample sizes. This shortcoming is balanced in part by the relatively small number of indicators required, and the use of sampling methods familiar to most field crews experienced in vegetation surveys. In addition, the ability to use existing vegetation data can reduce the number of new plots needed. Over time, the growth of repositories of vegetation inventory data will lower the number of new benchmark plots that need to be collected. Regulators should encourage companies to deposit samples of benchmark-quality vegetation in publicly available databases; companies themselves may voluntarily contribute such data in the interest of reducing future sampling burdens.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgments.\u003c/h2\u003e \u003cp\u003eWe thank Jared Hardner, Garry Luini, Katrina Lukianchuk, and Justin Straker and for suggestions that helped improve aspects of the VQA framework, and Christina Small for comments that strengthened the final manuscript. Dan Vasiga prepared the Elk Valley database exports of plots, vegetation and land cover used for the example application. Will Mackenzie and Deb MacKillop (BCENV) assisted with identification and selection of benchmark-quality plot data from the BCENV provincial vegetation database. This work was supported in part by funding from Teck Coal Limited.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBC Ministry of the Environment (2021) Terrestrial Ecosystem Data Standards \u0026amp; Guidelines. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www2.gov.bc.ca/gov/content/environment/plants-animals-ecosystems/ecosystems/tei-standards\u003c/span\u003e\u003cspan address=\"https://www2.gov.bc.ca/gov/content/environment/plants-animals-ecosystems/ecosystems/tei-standards\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBC Ministry of the Environment (2010) Field Manual for Describing Terrestrial Ecosystems, 2nd Editio. B.C. Ministry of Forests and Range, B.C. Ministry of Environment\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBC Ministry of the Environment (2016) BC Flora checklist 2016. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.for.gov.bc.ca/hre/becweb/resources/codes-standards/standards-species.html\u003c/span\u003e\u003cspan address=\"https://www.for.gov.bc.ca/hre/becweb/resources/codes-standards/standards-species.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 18 Jul 2016\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerger VW, Zhou Y (2014) Kolmogorov\u0026ndash;Smirnov Test: Overview. In: Wiley StatsRef: Statistics Reference Online. John Wiley \u0026amp; Sons, Ltd\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBest DJ, Gipps PG (1974) An Improved Gamma Approximation to the Negative Binomial. Technometrics 16:621\u0026ndash;624\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBritish Columbia Ministry of Forests-Lands and Natural Resource Operations (2016) BECMaster ecosystem plot database [MSAccess 2003 format]. Research Branch, Victoria, B.C. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.for.gov.bc.ca/hre/becweb/resources/information-requests/index.html\u003c/span\u003e\u003cspan address=\"http://www.for.gov.bc.ca/hre/becweb/resources/information-requests/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBritish Columbia Ministry of Forests-Lands and Natural Resource Operations (BCFLNRO) (2016) Biogeoclimatic Ecosystem Classification, Version 10\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCBD (2020) Update of the zero draft of the post-2020 global biodiversity framework\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCeballos G, Ehrlich PR, Barnosky AD, et al (2015) Accelerated modern human \u0026ndash; induced species losses: entering the sixth mass extinction. Sci Adv 1:1\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/sciadv.1400253\u003c/span\u003e\u003cspan address=\"10.1126/sciadv.1400253\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChazdon RL, Letcher SG, van Breugel M, et al (2007) Rates of change in tree communities of secondary Neotropical forests following major disturbances. Philos Trans R Soc Lond B Biol Sci 362:273\u0026ndash;89. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rstb.2006.1990\u003c/span\u003e\u003cspan address=\"10.1098/rstb.2006.1990\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCook CN, Wardell-Johnson G, Keatley M, et al (2010) Is what you see what you get? Visual vs. measured assessments of vegetation condition. J Appl Ecol 47:650\u0026ndash;661. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-2664.2010.01803.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2664.2010.01803.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCousineau D (2017) Varieties of confidence intervals. Adv Cogn Psychol 13:140\u0026ndash;155. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5709/acp-0214-z\u003c/span\u003e\u003cspan address=\"10.5709/acp-0214-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCumming G (2014) The New Statistics: Why and How. Psychol Sci 25:7\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0956797613504966\u003c/span\u003e\u003cspan address=\"10.1177/0956797613504966\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCurtin CG (1995) Can montane landscapes recover from human disturbance? Long-term evidence from disturbed subalpine communities. Biol Conserv 74:49\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0006-3207(95)00014-U\u003c/span\u003e\u003cspan address=\"10.1016/0006-3207(95)00014-U\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDemeo T, Haugo R, Ringo C, et al (2018) Expanding Our Understanding of Forest Structural Restoration Needs in the Pacific Northwest. Northwest Sci 92:18\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3955/046.092.0104\u003c/span\u003e\u003cspan address=\"10.3955/046.092.0104\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDukes JS, Mooney HA (2004) Disruption of ecosystem processes in western North America by invasive species. Rev Chil Hist Nat 77:411\u0026ndash;437\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEddy D, Hall R, Rehwinkel R, et al (2011) Assessing the assessors: Quantifying observer variation in vegetation and habitat assessment. 12:144\u0026ndash;148\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEfron B (1979) Bootstrap Methods: Another Look at the Jackknife. Ann Stat 7:1\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1214/aos/1176344552\u003c/span\u003e\u003cspan address=\"10.1214/aos/1176344552\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEyre TJ, Kelly A., Neldner VJ, et al (2011) BioCondition: A Condition Assessment Framework for Terrestrial Biodiversity in Queensland. Assessment Manual. Version 2.1. Brisbane\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranklin CW, Burton A (2018) End land use objective planning : integrating an ecosystem based approach into biodiversity and reclamation planning. In: 41st Annual TRCR Mine Reclamation Symposium. Fort Williams, BC, Canada, pp\u0026nbsp;1\u0026ndash;11\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFul\u0026eacute; PZ, Covington WW, Moore (1997) Determining reference conditions for ecosystem management of southwestern ponderosa pine forests. Ecol Appl 7:895\u0026ndash;908\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGastwirth JL (1975) Statistical measures of earnings differentials. Am Stat 29:32\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00031305.1975.10479109\u003c/span\u003e\u003cspan address=\"10.1080/00031305.1975.10479109\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGentry AH (1993) Patterns of diversity and floristic composition in neotropical montane forests. In: Churchill SP, Balslev H, Forero E, Luteyn J (eds) Proceedings of the Neotropical Montane Forest Biodiversity and Conservation Symposium, The New York Botanical Garden, 21\u0026ndash;26 June 1993. New York Botanical Garden, Bronx, N.Y., pp\u0026nbsp;103\u0026ndash;126\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGentry AH (1988) Changes in plant community diversity and floristic composition on environmental and geographical gradients. Ann Missouri Bot Gard 75:1\u0026ndash;34\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGentry AH (1995) Diversity and floristic composition of neotropical dry forests. In: Bullock SH, Mooney HA, Medina E (eds) Seasonally Dry Tropical Forests. Cambridge University Press, Cambridge, pp\u0026nbsp;146\u0026ndash;194\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibbons P, Ayers D, Seddon J, et al (2008) Biometric 2.0: A Terrestrial Biodiversity Assessment Tool for the NSW Native Vegetation Assessment Tool - Operational Manual. Canberra\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibbons P, Freudenberger D (2006) An overview of methods used to assess vegetation condition at the scale of the site. Ecol Manag Restor 7:S10\u0026ndash;S17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1442-8903.2006.00286.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1442-8903.2006.00286.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolmes P, Stuart-Smith K, MacKillop D, et al (2018) Old and Mature Forest Cumulative Effects Assessment Report: Elk Valley, Kootenay-Boundary Region\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHurlbert SH (1984) Pseudoreplication and the design of ecological experiments. Ecol Monogr 54:187\u0026ndash;211\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInternational Finance Corporation (2012) IFC Sustainability Framework: Policy and performance standards on environmental and social sustainability. Washington, DC, USA\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJennings MD, Faber-Langendoen D, Loucks OL, et al (2009) Standards for associations and alliances of the U.S. National Vegetation Classification. Ecol Monogr 79:173\u0026ndash;199. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1890/07-1804.1\u003c/span\u003e\u003cspan address=\"10.1890/07-1804.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKappelle M (1995) Ecology of mature and recovering Talamancan montane Quercus forests, Costa Rica. University of Amsterdam, Amsterdam\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKier G, Dinerstein E, Ricketts TH, et al (2005) Global patterns of plant diversity and floristic knowledge. J Biogeogr 32:1107\u0026ndash;1116\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnopff K, Franklin WE (2018) Biodiversity management: establishing pre-existing baseline conditions on mature and historical mining disturbances to derive back-casted wildlife habitat suitability model metrics for ten wildlife species. In: 41st Annual TRCR Mine Reclamation Symposium. Fort Williams, BC, Canada\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorb JE, Covington WW, Ful\u0026eacute; PZ (2003) Sampling Techniques Influence Understory Plant Trajectories After Restoration : An Example from Ponderosa Pine Restoration. 11:504\u0026ndash;515\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLandres PB, Morgan P, Swanson FJ, et al (2014) Overview of the Use of Natural Variability Concepts in Managing Ecological Systems. Ecol Appl 9:1179\u0026ndash;1188\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLewis M, Christianson A, Spinks M (2018) Return to flame: Reasons for burning in Lytton First Nation, British Columbia. J For 116:143\u0026ndash;150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/jofore/fvx007\u003c/span\u003e\u003cspan address=\"10.1093/jofore/fvx007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacKenzie W (2012) Biogeoclimatic ecosystem classification of non-forested ecosystems in British Columbia. Victoria, B.C.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMagurran AE (2004) Measuring biological diversity. Blackwell Publishing Ltd, Oxford, UK\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaron M, Hobbs RJ, Moilanen A, et al (2012) Faustian bargains? Restoration realities in the context of biodiversity offset policies. Biol Conserv 155:141\u0026ndash;148. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biocon.2012.06.003\u003c/span\u003e\u003cspan address=\"10.1016/j.biocon.2012.06.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCarthy M, Parris K (2004) The habitat hectares approach to vegetation assessment: An evaluation and suggestions for improvement. Ecol Manag Restor 5:24\u0026ndash;27\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcElhinny C, Gibbons P, Brack C, Bauhus J (2005) Forest and woodland stand structural complexity: Its definition and measurement. For Ecol Manage 218:1\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.foreco.2005.08.034\u003c/span\u003e\u003cspan address=\"10.1016/j.foreco.2005.08.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeidinger D V, Pojar J (1991) Ecosystems of British Columbia\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParkes D, Newell G (2003) Assessing the quality of native vegetation: the \u0026ldquo;habitat hectares\u0026rdquo; approach. Ecol Manag Restor 4:29\u0026ndash;38\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeet RK, Lee MT, Jennings MD, D. Faber-Langendoen D (2012) VegBank: a permanent, open-access archive for vegetation plot data. Biodivers Ecol 4:233\u0026ndash;241\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePimm SL, Joppa LN (2015) How Many Plant Species are There, Where are They, and at What Rate are They Going Extinct? Ann Missouri Bot Gard 100:170\u0026ndash;176. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3417/2012018\u003c/span\u003e\u003cspan address=\"10.3417/2012018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoole K, Teske I, Podrasky K, et al (2020) Bighorn Sheep Cumulative Effects Assessment Report: Elk Valley, Kootenay-Boundary Region\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRainey HJ, Pollard EHB, Dutson G, et al (2015) A review of corporate goals of No Net Loss and Net Positive Impact on biodiversity. Oryx 49:232\u0026ndash;238. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/S0030605313001476\u003c/span\u003e\u003cspan address=\"10.1017/S0030605313001476\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRio Tinto (2008) Rio Tinto and biodiversity: achieving results on the ground. Rio Tinto plc and Rio Tinto Limited\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRipley B, Venables B, Bates DM, et al (2013) Package \u0026lsquo;MASS\u0026rsquo;\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSahley CT, Vildoso B, Casaretto C, et al (2017) Quantifying impact reduction due to avoidance, minimization and restoration for a natural gas pipeline in the Peruvian Andes. Environ Impact Assess Rev 66:53\u0026ndash;65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.eiar.2017.06.003\u003c/span\u003e\u003cspan address=\"10.1016/j.eiar.2017.06.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchlesinger W, Andrews J (2000) Soil respiration and the global carbon cycle. Biogeochemistry 48(1):7\u0026ndash;20\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTerrestrial Ecosystem Mapping Alternatives Task Force Resource Inventory Committee (RIC) (1999) Towards the Establishment of Predictive Ecosystem Mapping Standards : A White Paper\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe State of Queensland (2014) BioCondition Benchmarks. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.qld.gov.au/environment/plants-animals/biodiversity/benchmarks/\u003c/span\u003e\u003cspan address=\"https://www.qld.gov.au/environment/plants-animals/biodiversity/benchmarks/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenables WN, Ripley BD (2002) Modern Applied Statistics with S, 4th ed. Springer Science \u0026amp; Business Media\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Environmental assessment, ecological monitoring, vegetation restoration, reclamation, loss-gain accounting, biodiversity offset","lastPublishedDoi":"10.21203/rs.3.rs-3374961/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3374961/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGovernments and society increasingly are demanding that industrial projects result in a net positive impact (NPI) on biodiversity. Impacts are commonly measured in terms of losses and gains of area and quality of vegetation, where quality refers to how closely a site matches the condition of native vegetation in its undisturbed state. Existing vegetation quality frameworks share a number of limitations, including little or no replication, uncertain scope of inference, vulnerability to bias, and inability to measure error. Here we present the Vegetation Quality Assessment (VQA) framework, a sampling-based extension of Quality Hectares that measures vegetation quality in terms of overlap between the probability distributions of ecological indicators at a project site and in undisturbed (benchmark) vegetation of the same kind. Distribution overlap incorporates natural variation at the landscape scale and provides an intuitive measure of quality that varies between 0 and 1. Indicators are measured using a stratified-random sampling design that minimizes bias and supports inference at the scale of the project landscape. Confidence limits of quality and quality hectares are determined by bootstrapping; power and minimum sample sizes are estimated by Monte Carlo simulation. Multiple assessments track losses and gains of quality hectares and enable accurate accounting of progress to NPI. The VQA framework can be implemented using a variety of vegetation sampling methods, allowing existing vegetation databases to be leveraged as sources of data. We conclude by demonstrating the application of VQA at several mining operations in the Elk Valley of southeastern British Columbia, Canada.\u003c/p\u003e","manuscriptTitle":"Vegetation Quality Assessment: a sampling-based, loss-gain accounting framework for native, disturbed and reclaimed vegetation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2023-09-28 15:50:12","doi":"10.21203/rs.3.rs-3374961/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ee147a37-d242-42d0-bab1-41837837d532","owner":[],"postedDate":"September 28th, 2023","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-03-27T06:18:08+00:00","versionOfRecord":{"articleIdentity":"rs-3374961","link":"https://doi.org/10.1016/j.ecolind.2023.111510","journal":{"identity":"ecological-indicators","isVorOnly":true,"title":"Ecological Indicators"},"publishedOn":"2024-01-01 06:18:08","publishedOnDateReadable":"January 1st, 2024"},"versionCreatedAt":"2023-09-28 15:50:12","video":"","vorDoi":"10.1016/j.ecolind.2023.111510","vorDoiUrl":"https://doi.org/10.1016/j.ecolind.2023.111510","workflowStages":[]},"version":"v1","identity":"rs-3374961","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3374961","identity":"rs-3374961","version":["v1"]},"buildId":"cBFmMYwuxLRRLfASyISRj","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}