The concept of persistence is embedded in the conservation goal of ecoregional assessments and biodiversity visions. The intention is to conserve native species, communities and ecological systems that will persist over time. The potential for persistence can be evaluated through assessing the viability of populations, and the ecological or biological integrity of communities and ecological systems. Viable populations maintain their rigor and potential for evolutionary adaptation (Soule 1987). Biological systems with integrity have the capacity to support and maintain the full natural range of biological elements and ecological processes (Karr and Chu 1995). Viable populations and communities and ecological systems with integrity can withstand and recover from natural and human-caused perturbations.

Assessing the viability or integrity of biodiversity targets and features is important for several reasons. Ecoregional portfolios and biodiversity visions identify areas of biodiversity significance that if conserved, will maintain the biodiversity of an ecoregion. It is critical that these areas are composed of populations, communities and ecological systems that are the best candidates for persistence, and they have an inherent capacity to maintain viability and integrity. Conserving areas with inherently persistent biodiversity is less costly and less risky that investing in restoration. Secondly, conservation goals suggesting the abundance and distribution of species, community and ecological systems needed to maintain them across an ecoregion, are dependent on having persistent examples in order to achieve those goals. Knowing the gap in achieving a conservation goal informs the need to inventory and identify more persistent examples of biodiversity, and to restore examples that are not currently viable or exhibiting integrity. Lastly, but not least, measuring progress in conservation is dependent on knowing the distribution and abundance of examples of biodiversity targets that are viable or exhibit integrity, and the degree of progress towards meeting goals.

Populations, especially small ones, are subject to great influence by chance events. Demographic uncertainty, random events in the survival and reproduction of individuals, and genetic drift can have significant impacts on the viability of populations. The persistence of populations, communities and ecological systems are all influenced by environmental uncertainty - unpredictable events related to weather, predators and competitors, as well as natural catastrophes and disturbance events, such as hurricanes, fires and floods. Characteristics that help them to persist in the face of these events include their size, condition, and their landscape context.

No area should be included in an ecoregional portfolio or biodiversity vision unless its targets (at a minimum, the coarse-filter targets) have been screened for viability or ecological/biological integrity. While many of the methods presented below provide several categories that describe ranges of viability, at a minimum, a viable/non-viable classification should be completed. Below we describe approaches that are applicable for:

• both viability and integrity assessments,
• assessing integrity, and
• assessing viability.

Approaches to Viability and Integrity Assessments

There are a number of approaches to assess the viability and integrity of populations, communities and ecological systems. Our principal recommendation is for ecoregional planners to work with experts and apply the criteria of size, condition, and landscape context (described below) to as many occurrences of conservation targets as is possible and practical. As a first priority, we strongly encourage planning teams to develop viability specifications for ecological systems. Next in importance is for teams to assess the viability of finer-scale community and species targets.

A fairly rapid approach to define population viability and community and ecological system integrity is to use single or combined criteria (depending on available data) of size, condition and landscape context (Stein and Davis 2000):

• Size is the abundance/density of a population, or the area of a population or ecological system.
• Condition is the quality of biotic and abiotic factors, structures and processes within a population or ecological system occurrence, such as age structure, species composition, ecological processes and physical/chemical factors.
• Landscape context is the quality of structures, processes and biotic/abiotic factors of the landscape surrounding a population or ecological system, including degrees of connectivity and isolation to adjacent habitats, populations and ecological systems.

Criteria 1 - Size. At the population level, size is a measure of the area of occupancy by a species and/or its population abundance and density. All else being equal, larger populations are assumed to be more viable than smaller populations. For matrix-type communities and ecological systems, large-scale natural disturbances create a diverse shifting mosaic of successional stages and physical settings. The necessary area needed to ensure survival or re-colonization from such disturbances (e.g., disease, fire, insect outbreaks, hurricanes) has been called the minimum dynamic area. For a matrix type to persist over time it must be able to sustain, buffer, and absorb these disturbances and maintain these minimum dynamic areas. Size can be determined in two ways for ecological communities and systems. First, the home range of a species (usually a vertebrate) that is a typical occupant of that system and is at the higher end of the food chain can be used to estimate the size of the community or system (e.g., Flammulated Owl in ponderosa pine forests). Alternatively, there is a rule of thumb from the field of patch dynamics and disturbance ecology that suggests the size of a community or system needs to be the size of the largest natural disturbance to that community or system over a 500-1000 year time frame.

For aquatic communities and systems, large-scale natural disturbances like floods and droughts create a mosaic of habitat suitability. Aquatic organisms will often move to refugia during disturbance events and recolonize after habitat conditions become favorable again. A minimum dynamic area for aquatic systems must be large enough to ensure the linear connectivity of habitats at scales appropriate to the targets. As with populations, larger occurrences for communities and systems are generally preferable to smaller ones, especially for matrix types.

Criteria 2 - Condition. Condition is an integrated measure of the quality of biotic and abiotic factors, structures, and processes that characterize targets. Criteria for measuring condition include success and regularity of reproduction, presence/absence of competitors/predators, degree of anthropogenic impacts and presence of biological legacies.

• Anthropogenic impacts - fragmentation, presence of exotic species, alteration of natural disturbance regimes, pollution, and so on. Occurrences that contain relatively continuous cover of natural vegetation (i.e., less fragmentation) are more likely to have intact ecological processes and be free of invasive exotic species.
• Biological legacies - critical features of communities and systems that take generations or sometimes hundreds to thousands of years to develop. For example, in old-growth forests the presence of fallen logs and rotting wood, a well-developed herbaceous under-story, and structural complexity in the canopy are examples of biological legacies. As a general rule, the presence of a well-developed structure and species composition that include characteristic, but also uncommon species, implies good habitat quality and some historical continuity. Those communities and systems that are depauperate in species composition for any of a variety of reasons make poor "coarse filters" In regions with high species diversity, but are important in regions that are naturally low in species diversity.

Criteria 3 - Landscape Context. For populations, landscape context is an integrated measure of two criteria: connectivity to other populations and intactness of surrounding ecological processes and environmental regimes. Although landscape context is important for all communities and systems, those patch and matrix types and aquatic communities and systems that depend on easily disrupted ecological processes occurring at a scale larger than the individual community are most at risk by what happens in the surrounding landscape (e.g., altered fire regime, altered flow regime, ground water pumping). A few patch communities such as those on raised bogs, perched wetlands, isolated lakes, and cliffs and rocky summits are more dependent upon atmospheric input of nutrients and water than the surrounding landscape. In general, communities and systems that are connected to or in proximity to other natural habitats are usually preferable to isolated examples.

In the United States and parts of Latin America, Natural Heritage Programs, Conservation Data Centers and NatureServe have developed and applied element occurrence ranks based on these criteria for many species and communities, and are developing them for ecological systems. Ranks of A (excellent estimated viability), B (good), C (fair), and D (poor) are available in their databases. Where these are available, they should be applied. Occurrences with a rating of Poor (D) should not be considered viable, and any Fair (C) ratings should be accepted with some caution. Such data will largely be available only for communities (i.e., plant associations, not ecological systems) and usually only for highly ranked (G1-G2) communities. If resources allow, expert opinion or site visits should be used to assess viability of community occurrences for which no information is available. Alternatively, GIS analytical approaches may be used to readily assess size and landscape context.

Specific Integrity Approaches

Approaches to evaluate ecological integrity include assessing the potential for communities and ecosystems to be composed largely of their native species, and have ecological processes and natural disturbances such as fires and floods operating within their natural range of variation (Groves 2003). Some principle criteria used to assess community and ecosystem viability is minimum dynamic area (Pickett and Thompson 1978) and natural flow regimes (e.g. Poff et al. 1997). In the United States, the Index of Biotic Integrity is a widely applied measure of freshwater ecosystems that can be obtained from state and federal Environmental Protection Agencies.

The most rapid method to assess community and ecological system integrity is to develop suitability indices using spatial information and other available data. These indices provide a way to evaluate the relative conditions and landscape contexts of spatial units such as ecological systems, watersheds, or comprehensive regular mapping polygons such as hexagons. The indices are built from combining information from several data layers. Data layers commonly used for terrestrial and freshwater indices include: number of sensitive species, percent natural land cover, road density, human population density, degree of fragmentation, and community/ecological system size (e.g. Moyle and Randall, 1998, Davis et al. 1999. See Groves 2003 for other examples). A readily available global data layer to assess anthropogenic impacts is the Human Footprint (see resources) which was developed using data indicating the extent of human population pressure, land use, infrastructure and access (Sanderson et al. 2002). Developing a suitability index in itself does not define thresholds for viability. Often, spatial patterns are categorized by quartiles or quintiles, and suitability indices identify clearly those targets that are affected the most by many factors, and are probably not viable, and those that are affected the least, and have the highest relative integrity. However, this does not ensure integrity. It merely provides an initial screening to identify those examples that are the best candidates for inclusion in an ecoregional portfolio or biodiversity vision because they have the best size, and are in the best condition and landscape context. Working with regional experts to define thresholds for integrity is strongly urged.

Viability Specific Approaches

Population viability analysis (PVA) is the quantitative assessment and modeling of the probability of a population to go extinct. What is most helpful to conservation planners is the minimum population size necessary to maintain viability. In addition to population size, understanding meta-population dynamics and the landscape context of single, large populations vs. many small, populations with immigration and emigration among them is helpful. Unfortunately, few species have had PVAs conducted on them, and most of them that have are in relatively data-rich areas. Groves (2003) suggests that PVAs are most helpful for site-based management planning, and generally not very helpful for regional assessments. However, if a planning team is interested in using PVA or other population viability assessments, they should consult available sources. In the United States, many PVAs have been applied to species listed under the Endangered Species Act, and there are recovery plans for listed species (see resources). Practitioners working internationally may find it useful to consult IUCN Action Plans (see resources) for endangered, critically endangered, and vulnerable species in order to assess the viability of target species' populations. These plans typically include a Population and Habitat Viability Analysis (PHVAs). This tool, developed by the IUCN Conservation Breeding Specialist Group, focuses on specific factors affecting the status of the population and recommends conservation action.

In the absence of rigorous population viability analysis, conservation planners have used rules of thumb: defining how many individuals are needed to maintain a population in the face of demographic, genetic and environmental uncertainties, as well as natural and human caused catastrophes. Recent analyses suggest a census population of 5000 individuals for many species (Allendorf and Ryman 2002). This is often unachievable. However, there are other approaches to maintaining species with lower population sizes including improving landscape context through increasing connectivity among geographically isolated populations (Groves 2003).

Empirical habitat suitability models have been developed for a variety of species, such as wide ranging carnivores (e.g. Carroll et al. 2001), birds and mammals (e.g. Block et al. 1994), and freshwater fishes (e.g. Filipe et al 2004). These approaches can indirectly provide an indication of the potential for viable populations in a given habitat.

Key Steps

• Identify and use existing information on population viability analyses and management recovery plans for species, and viability/integrity rankings for communities and ecological systems. Develop criteria for acceptable levels of viability (e.g. A and B ranks from Natural Heritage Programs, C ranks with justifications), and for acceptable age of data (e.g. do not apply any ranks that are older than 20 years in a landscape that is undergoing measurable change).
• Identify available survey and spatial data, and methods to fill in gaps in viability/integrity assessments.
• Apply a method using GIS data and technology to evaluate size, condition and landscape context. Develop empirical models of spatial patterns and on-the-ground ranks to inform ranking criteria and thresholds (if possible).
• Work with regional experts to define criteria and ranks for viability/integrity using available information and spatial data in the framework of size, condition and landscape context. At a minimum, define viable/non-viable categories.
• Document data sources, procedures and decisions used to define viability/integrity.
• Document data gaps and next steps to fill gaps and to strengthen future assessments
• Document the number and distribution of target occurrences (examples) that are viable or have integrity, and measure the progress towards goals for these targets. Summarize using guidance in the Measures standard.

NOTE: If a suite of targets do not have any occurrences that meet minimum viability/integrity criteria, the best examples can be included in a portfolio, but they cannot be included in progress towards achieving goals. They will be initial restoration priorities.