Mind the GAP—But make it better: Improving the U.S. Gap Analysis Project's protected‐area classification system to better reflect biodiversity conservation

Protected areas are foundational to the conservation of biological diversity, and many scientists have called for increased protections in the face of the current extinction and climate‐change crises. Currently in the United States, the most recognized way to track the amount of protected land is the Gap Analysis Project classification system, which we argue is deficient in three ways: it does not, in a systematic way, specify the typical uses and constraints associated with each conservation designation; it is not fine‐tuned or nuanced enough to distinguish the levels of protection among designations within “protected” or “unprotected” areas that allow quite different human activities; and it largely ignores the durability of the designations, failing to account for an area's vulnerability to downsizing, downgrading, or degazetting. We propose a new classification system to address these deficiencies and demonstrate this method for several of the most common federal land designations in the United States.


| INTRODUCTION
Protected areas are the foundation of biodiversity conservation Bingham et al., 2021) because they reduce the loss and degradation of natural habitats (Bruner et al., 2001;Naughton-Treves et al., 2005) and slow the rate of extinction of threatened species (Butchart et al., 2012). Although protected areas are often insufficient by themselves to conserve biodiversity (Hayward, 2011), they provide a necessary anchor to the complex management actions necessary to sustain nature. Studies have shown that biodiversity is substantially higher inside than outside well-managed terrestrial reserves (Gray et al., 2016) because human-caused disruptors of habitat quality such as roadbuilding, mining, oil and gas drilling, commercial logging, livestock grazing, hunting, or motor-vehicle recreation are often restricted inside their boundaries.
Therefore, scientists, conservation organizations, and governments around the globe have repeatedly called for protecting more lands and waters in nature reserves to ensure that biodiversity and natural processes are conserved. While scientists differ with respect to precise areabased targets, analyses from empirical data and models suggest that "25%-75% of a typical region must be managed with conservation of nature as a primary objective to meet goals for conserving biodiversity" (Noss et al., 2012, p. 2). Based on the precautionary principle, 50% of land protected in such reserves-slightly above the midpoint of recent evidence-based estimates-has been called "scientifically defensible as a global target" (Noss et al., 2012, p. 2;Wilson, 2016). As an interim measure, many policymakers have adopted a goal of protecting 30% of lands and waters in every major ecosystem type by 2030 (i.e., "30 Â 30").
The concept of 30 Â 30 originated from an international team of scientists who proposed a Global Deal for Nature, whereby 30% of the Earth would be formally protected by 2030 to prevent mass-extinction of vulnerable species (Dinerstein et al., 2017(Dinerstein et al., , 2019. As part of this effort, American President Joseph R. Biden Jr. issued, in 2021, Executive Order 14008, which directs the U.S. Department of Interior to establish mechanisms to measure progress toward the goal of conserving at least 30% of U.S. lands and waters by 2030 (E.O. 14008 § 216[a]ii).
This prompts two questions: what is a protected area and why is it classified as such? The two primary inventories of protected areas used in the United States are the World Database on Protected Areas (WDPA; UNEP, 2022) and the U.S. Geological Survey's Gap Analysis Project (GAP) Protected Areas Database (PAD-US; USGS, 2020). PAD-US is "America's official national inventory of U.S. terrestrial and marine protected areas that are dedicated to the preservation of biological diversity and to other natural, recreational, and cultural uses, managed for these purposes through legal or other effective means." 1 Currently in the United States, the most recognized way to categorize the degree of protection of different conservation designations in the PAD-US is the GAP classification system of protected areas. The GAP classification system relies on broad definitions that reflect management intent with respect to biodiversity conservation. GAP status codes of 1, 2, 3, or 4 are assigned to land designations to reflect management intent to conserve biodiversity (Table 1). When translating the GAP system to the system used by the WDPA, 2 GAP status 1 and 2 areas are recognized by the United Nations Environment Program and the International Union for Conservation of Nature (IUCN) as "Protected Areas." GAP statuses 3 and 4 are not (UNEP, 2022). The question remains whether this system is ideal for tracking progress toward biodiversity conservation.
In our experience, some land managers and conservationists have expressed concerns about the current GAP classification system because of its implied dichotomy (i.e., "protected" or "unprotected") or its overly broad classifications within "protected" or "unprotected" areas. Despite the criticism, this broad and dichotomous protected-area classification has been adopted in several national conservation assessments (Aycrigg et al., 2013;Belote et al., 2017;Dietz et al., 2020;Lawler et al., 2020).
We believe, however, that the current American system for classifying protected areas and tracking our

GAP status domain code
Domain Description 1 Managed for biodiversity-disturbance events proceed or are mimicked An area having permanent protection from conversion of natural land cover and a mandated management plan in operation to maintain a natural state within which disturbance events (of natural type, frequency, intensity, and legacy) are permitted to proceed without interference or are mimicked through management.
2 Managed for biodiversity-disturbance events suppressed An area having permanent protection from conversion of natural land cover and a mandated management plan in operation to maintain a primarily natural state, but which may receive uses or management practices that degrade the quality of existing natural communities, including suppression of natural disturbance. 3 Managed for multiple uses-subject to extractive (e.g., mining, logging) or OHV use An area having permanent protection from conversion of natural land cover for the majority of the area, but subject to extractive uses of either a broad, low-intensity type (e.g., logging, off-highway vehicle recreation) or localized intense type (e.g., mining). It also confers protection to federally listed endangered and threatened species throughout the area.

No known mandate for biodiversity protection
There are no known public or private institutional mandates or legally recognized easements or deed restrictions held by the managing entity to prevent conversion of natural habitat types to anthropogenic habitat types. The area generally allows conversion to unnatural land cover throughout or management intent is unknown.
Note: The GAP status code is a measure of management intent to conserve biodiversity.
progress toward area-based targets could be improved by addressing three deficiencies. First, the current system does not, in a systematic way, specify the typical uses and constraints associated with each conservation designation (Table 1). Therefore, the system does not enumerate the actual habitat disruptors-the mechanisms-that will most likely predict success or failure of an area in conserving species of conservation concern. Second, the system is not fine-tuned or nuanced enough to distinguish the levels of protection among designations within "protected" or "unprotected" areas that allow quite different human activities. As a result, the system conceals important differences among lands classified by the same numeric code ( Figure S1). Third, the current classification system largely ignores the durability of the designations, failing to account for an area's vulnerability to downsizing, downgrading, or degazetting (see Golden Kroner et al., 2019).
Here, we recommend a revised method for classification of protected areas that specifies typical uses and constraints, enumerates and weighs the specific habitat disruptors that may threaten species of conservation concern, weighs the durability of each designation, and expands the classification codes by adding subclasses that better reflect differences in levels of protection. We demonstrate this method for several of the most common federal land designations in the United States, although the process is designed to work across all ownership and management types.

| METHODS AND RESULTS
We propose a simple, four-part approach to improving the GAP classification system on all U.S. lands: 1. Retain the four GAP status domain codes and definitions (from Table 1). 2. List the major categories of habitat disruption (Table 2, first column) that threaten vulnerable species and quantify their allowance or restriction, through weighted scores, in each designation type. 3. Identify the law, regulation, or policy behind each designation type and assign a weighted "durability" score for each. 4. Sum the habitat disruption score and the durability score to assign a new GAP code consisting of a numeric (1-4) and letter (a-c) code in a hierarchical classification system that better characterizes protection within an area's boundaries (Table 2, second to last row).
The approach described here builds on concepts introduced in Aplet and McKinley's (2017) Wilcove et al. (1998) and additional exploitation factors (hunting, fishing, plant gathering). It enumerates which of these habitat disruptors typically occurs within a designation type based on the laws, regulations, and policies associated with each designation. For demonstration purposes, we weighted each disruptor according to our assumed level of threat to vulnerable species. This classification system provides a good indication of how well protected from habitat disruptors an area would be under each designation type.
We graded the durability of protection for a given area of federal land as high (weighted score = 0), medium (5), or low (10) based on the degree of difficulty required to reduce protections. We applied a high durability ranking to areas protected by law, which includes areas specifically designated by an Act of Congress and the signature of the President, such as national parks and wilderness areas, as well as areas designated using authorities granted by Congress and delegated to the Executive Branch, such as national monuments designated by the president, national wildlife refuges created by the Secretary of the Interior, and wilderness study areas designated by the Bureau of Land Management (BLM). Medium durability is assigned to areas protected through regulations developed by federal agencies under authority granted by law and using the process required by the Administrative Procedure Act (i.e., Federal rulemaking). An example of medium durability is the designation of U.S. Forest Service Inventoried Roadless Areas (IRAs), which are managed according to the Roadless Area Conservation Rule (36 Code of Federal Regulations § 294). A low durability score is given to areas where protections can be applied or removed by agencies using their designating authority, often through or in concurrence with approval of a management plan amendment or revision. This would include designations such as Areas of Critical Environmental Concern (ACEC), experimental forests, research natural areas, and botanical areas.
The weighted scores for habitat disruptors and durability of the designation are then summed to reach a composite score that determines the expanded GAP code. We added hierarchical subclasses (1a, 1b, 2a, 2b, 3a, 3b, 3c, 4a, 4b) to this system to acknowledge the value in accounting for a broader and more nuanced continuum of protection. The results of this scoring system are shown in the second-to-last row of Table 2.

| DISCUSSION
The overarching goal of the GAP classification system is to determine where biodiversity conservation lands are, who manages them, and how much of a plant or animal species' habitat is included in various management types-i.e., where the conservation "gaps" are. 3 Assigning the level of protection for land designations is paramount to the effectiveness of the program in conserving biodiversity. Currently, the classification system relies on broad domains that describe whether and how an area is managed for biodiversity (Table 1). We have no issue with the domain definitions or the labeling of GAP 1 and 2 lands as "protected" and GAP 3 and 4 lands as "unprotected." We do, however, think that the system could be improved by enumerating the specific types of potential habitat disruptors (and exploitation, such as hunting) typically allowed or constrained in each type of designation and by assigning a score and subclass accordingly. This proposed method provides a better defense for the protected/unprotected dichotomy because it is clearer and more transparent about the criteria used to make those decisions.
Why are habitat disruptors the most appropriate mechanisms with which to rank protected areas? For three reasons. First, they are the most pervasive cause of imperilment among plants and animals in the United States and globally (Di Marco et al., 2018;Harfoot et al., 2021). Habitat degradation and loss was shown to be the cause of imperilment of 85% of plant and animal species (n = 1880) in the United States, compared to 49% for introduction of alien species, 24% for pollution, 17% for overexploitation, and 3% for exotic diseases (Wilcove et al., 1998). Second, they are the factors that are typically allowed or constrained in a designated area-as opposed to pollution, introduction of non-native species, or introduction of exotic diseases, which are typically not regulated by management type. Third, habitat disruptors are often the direct or indirect cause of pollution, introduction of non-native species, or introduction of exotic diseases, irrespective of intention or regulation. Exploitation of biological resources-mostly in the form of hunting, fishing, and plant gathering-are typically regulated by management type. Thus, they are included in our list of "habitat disruptors," even though they are often considered a separate category.
Currently, dozens of federal, state, and local designation types with diverse management practices and use restrictions are grouped into only four domains. Lumping many designation types into four domains conceals important details and differences in conservation-area management and policy. Quantifying habitat disruptors is a necessary step in revealing the true differences among conservation designations. For instance, federal lands classified as GAP status 3 include IRAs, ACECs, National Recreation Areas, and general U.S. Forest Service and BLM lands open to timber harvesting, mining, and many other extractive or disruptive uses (Table 2). An excessively wide range of use restrictions and management practices are grouped into this one domain. Similar issues exist with GAP statuses 1 and 2. For instance, designated wilderness areas within national parks are classified the same as wilderness areas in national forests (GAP 1). The former prohibits hunting animals and gathering plants, whereas the latter allows state-regulated hunting and plant gathering. This distinction can easily be accounted for by adding, for example, subclasses 1a (national park wilderness) and 1b (other wilderness). Currently, a single habitat disruptor-fire suppression-distinguishes between two GAP codes (1 and 2), while GAP 3 areas contain any number of other habitat disruptors. In addition, GAP 4 does not distinguish between a known lack of mandate for biodiversity conservation and a lack of knowledge if there is a mandate. Our proposed system differentiates between the two. Classifying anything necessitates deciding whether to "lump" or "split" items, and each approach entails tradeoffs. We feel there needs to be a better balance between lumping and splitting by adding subclasses within domains. This approach retains the current "protected" and "unprotected" dichotomy (lumping)-which is necessary for determining area-based protection targets-while expanding the continuum of levels of protection (splitting), which allows for better tracking of degrees of protection.
With respect to assessing durability, the current GAP classification system is deficient. In the three GAP domains where there is a known mandate that prevents conversion of natural land cover, there is no definitional distinction among the domains regarding durability. They each assume "permanent protection." 4 But there are, in fact, degrees of durability among conservation designations. In our examples, we have assessed the durability of conservation designations on federal lands of the United States and grouped them into three categories. Clearly, no conservation designation is truly permanent, but we believe it is useful to gauge the difficulty for a governing body to downgrade or downsize a protected area (Golden Kroner et al., 2019).
What we have outlined here is a conceptual model for an improved method to classify protected areas in the United States. We have made choices in this model that others may not agree with. For instance, one could assign different weights to each habitat disruptor-as we didor simply tally the number of disruptors typically allowed in each designation type. Many scientists believe that ranking specific threats to biodiversity is context dependent (Bellard et al., 2022) and therefore fraught. In our examples, our weighted method and an unweighted method resulted in highly correlated disruptor scores (r 2 = .90). Some may have a different opinion about whether a particular habitat disruptor is "typically and commonly" allowed in each designation. In fact, several conservation designations do lack specificity about the uses and constraints that we use in our assessment. For example, national monuments are generally considered to be "protected" areas, yet the use restrictions in a monument are largely the product of the management plan that is unique to that unit. It is important to note, therefore, that the GAP status codes in Table 2 are "default" classifications that may be adjusted for an individual unit after a more in-depth evaluation including local and national experts. The important thing to take away from this model is the conceptual framework for classifying any land designation based on the habitat disruptors and durability, irrespective of ownership, manager, or size.
Adding subclasses to the GAP system responds directly to the Biden Administration's call to "reflect a continuum of conservation actions" (Fed. Reg., Vol. 87, No. 2) in tracking America's conservation progress. The current system's "continuum" is limited to four ordinal classes; our proposal would better represent a gradient-or continuum-of conservation protections. This new system would provide a more complete, detailed, and accurate reflection of complex management and policy providing hierarchy to the current system. Users can choose to focus on the original GAP 1-4 domains or concentrate on the finer distinctions among subclasses. One might argue that the IUCN classification, which includes seven classes, could be a reasonable alternative to adopt to better represent a conservation continuum. However, IUCN classes are categorical and not necessarily ordinal (e.g., IUCN category III is not necessarily better protected from disruptors than category IV). Just as the current PAD-US "crosswalks" to IUCN categories, our proposed revised classification would include a similar translation.
Specifically, the benefits of this new classification system are many: it provides a) information on the specific uses and constraints in an area, rather than an overly broad domain description only, b) the opportunity to create a simple online form for managers that would require specific detail about management intent, c) a more complete accounting of conservation promotions and "upgrading" which would allow federal, state, and private agencies and institutions to better track-and get credit for-more modest but important shifts in management and policy without "watering down" requirements for formal protected areas, d) greater incentives for elevating area-based protections, as the expanded continuum would recognize improvements outside of the protected/unprotected dichotomy or the limits of overly broad GAP domains, e) increased transparency about and responsibility toward taking management actions that reflect the specific uses and constraints that compose the habitat-disruptor score, f) greater incentives for land managers to submit detailed data about conservation lands, and g) a methodology that makes explicit that all natural-area conservation actions for biodiversity matter-irrespective of whether they "count" for the 30 Â 30 biodiversity goal.

DATA AVAILABILITY STATEMENT
All information is included in the article.