Protocols and Models for Inventory and Monitoring of Threatened and Endangered Plant Species: Part 1, continued; Protocol for Inventory and Monitoring of Threatened and Endangered Plant Species

Part 1 Continued: Sections IV & V
[Return to Sections I-III]
Protocol for Inventory
and Monitoring of Threatened
and Endangered Species

The preliminary survey, for the purposes of this Protocol, is the process that determines which TES may be present in an administrative planning unit. The preliminary survey involves the use of local experts, consultation with USFWS, searches of the literature and searches of information databases, to collect available information that can be used to guide or expedite further activities involving TES. Information may include: lists of TES most likely to occur in the area, existing knowledge of the community types represented in the area, records of disturbance patterns, and compilations of area maps and other relevant graphical materials.

Identify TES that occur or may occur within the administrative planning unit (APU).

The purpose of the preliminary survey is to identify TES that may occur in an area where management is planned. In some cases, consultation with USFWS or results from previous work may provide part or all of the information required for the preliminary and field survey phases. For this reason, a thorough literature review and inquiry among local experts should be conducted to eliminate unnecessary redundancy of effort. What follows in this subsection presumes that the presence and distribution of TES are not known. Where previous surveys are available, NR managers may alter procedures accordingly.

Information about potential TES and their habitat can be supplemented through reference to surveys and inventories of adjacent or similar areas within the known range of the species. Other sources of information on possible TES occurrence include databases maintained by state Natural Heritage Programs and local chapters of The Nature Conservancy; state conservation departments; area universities; state native plant societies; federal agency personnel working in rare plant conservation; regional herbaria and their personnel; and published sources such as area floras, plant manuals, and academic journals and databases. No preliminary survey can canvass all, or even a majority, of these sources. Personal contact with knowledgeable workers associated with some of them will often provide a synthesis of information from sources not actually surveyed.

Gather species and identification information for potentially occurring TES from the literature and from experts. Use existing species information to identify special requirements, associations, or sensitivities of TES and to refine installation or public land habitat information.

Thorough taxonomic knowledge of each TES will be necessary to provide background and taxonomic information for workers who conduct field searches. TES may be similar in appearance or closely related to common plant species, which may complicate identifications. Species descriptions, photographs and illustrations can be compiled for field use in order to facilitate identification.

A.2. Collection of Existing Habitat and Land-Use Information About the Management Area

Use available habitat information to delineate areas of possible habitat for TES plants that may occur on the management area.

Primary clues as to which species may be present is provided by knowledge of the range of habitats represented on the area under consideration. At early stages of assessment, detailed habitat information may not be available; however, the NR manager may assemble qualitative habitat information by noting the presence of broad vegetation types such as "oak-hickory forest," "scrub desert," "grassland," or riparian areas, as well as more restricted and specific habitats like areas of unusual soil types or geology, restricted plant communities, disturbed areas, and the like.

Habitat information can also be obtained from existing aerial and satellite images; soil, watershed, vegetation, and topographic maps; and other available graphical resources that describe relevant natural features of the area of interest. Recent advances in the use of airborne video systems show promise as an efficient method for obtaining digital habitat information (Myhre et al. 1991). Even limited habitat information can narrow the TES list considerably. For example, species with regional distributions, but that are restricted to scattered patches of particular kinds of forests would not be expected to occur on an installation or other area that does not include such a forested habitat, which should be evident from aerial photography or videography. See Subsection IV.A.3 for further discussion of the use of habitat information for the prediction of TES occurrence.

Review past, current, and projected activities on the public land area or installation.

Knowledge of land uses that impact natural communities will be needed to identify potentially detrimental effects of land-use practices on TES populations. Examples of land use activities include revegetation projects, road building, development projects, and water diversions. Some activities, for example, shelling or equipment fueling, may release contaminants detrimental to TES and their habitats. Disturbance activities may have caused retrogression of plant communities to early secondary succession. Knowledge of the location of these activities can provide valuable and timely information for species dependent on or sensitive to disturbance. Drastically altered habitat can, in most cases, be eliminated from searches.

Managers should develop a preliminary list of TES plants that may occur if it is not known which TES actually occur. The list of TES judged to have potential for occurrence will be essential to the design of the field survey (Subsection IV.B). Where habitat information is lacking, the list should include species judged to have low probabilities of occurrence since the list will be used to develop specific objectives of the field personnel engaged in searches to verify TES occurrence.

Various predictive methods can be used to facilitate location of TES populations. Predictions should be verified by field surveys.

Predictive approaches, whether qualitative, geographic, or quantitative, may be used to facilitate TES surveys through identification of those habitats that most likely contain TES populations. The species must have distinct habitat requirements for predictive approaches to be effective, this information must be known for the species, and habitat information must be adequate for the area. Where this information is available, modeling can assist in reduction of the area searched.
Map overlays (done manually or using GIS technology) of known TES locations may simplify the location of populations on an unsurveyed area. Nelson (1985) described an example in which known populations of Polygonum bidwelliae (endemic to Butte County, CA) were located on topographic, geological, and vegetation maps, which provided limiting information on elevation, substrate, and plant communities that characterized the habitat of this species. Map overlays allowed potential locations of unknown populations to be inferred wherever these three conditions were met, but populations had not yet been recorded.

Surveys may be supplemented by various quantitative predictive methods such as statistical models. Regression and correlation analyses can be used to predict species occurrence given the occurrence of a set of habitat components found to be correlated with the presence of the species. Spatial models using autoregressive techniques enable a study of the relationship of a species and its environment over an area (Bonham, et al. 1995, integrated into Part II). Canonical correlation methods can also be used to model plant-environmental (habitat) relationships (Green 1979b).
Variants of the predictive modeling techniques that have been used recently by federal agencies to model the distribution and abundance of animals may have applications for prediction of plant TES presence. An unpublished 1995 report on habitat characterization by Bonham et al. (1995a, integrated as Part II ) reviews models used by ecologists to describe relationships between species and their environments. It provides suggested strengths, limitations, and potential uses of specific model types for plant TES. In animal studies, reference is usually made to "habitat" descriptions while in plant studies the term "environmental factor" is more often used. The Protocol does not distinguish between these terms.

For example, habitat suitability index (HSI) models and PATREC models, both traditionally used for wildlife, may have applications for locating plant TES populations in defined geographic areas if sufficient habitat information is available at scales suitable for the TES in question. An HSI is a numerical value that represents the capacity of a given habitat to support a selected species (USFWS 1981). The HSI is often based on an assumed linear relationship between several habitat variables and population density. The range of variation in habitat variables for a species is assigned suitability index (SI) values on a scale of 0 to 1. These indices are used to estimate HSI values to predict potential habitat for a species. HSI models have not been found to be highly predictive for wildlife species.

A PATREC model uses a set of habitat attributes that is assigned conditional probabilities, based on field observations, to predict habitat potential for a species. A conditional probability is assigned for high and low population densities observed under various habitat conditions (Smith et al. 1991). Bayes' theorem (see Williams et al. 1978) is used to calculate a posterior probability that the area will support a high-density population.

More traditional, vegetation-oriented techniques, such as ordination (ordering of species values with respect to similarity of groupings or environmental values) may also be used to identify the plant communities with which a TES is most likely to be associated (e.g., Bowles et al. 1993). Such methods may help to narrow search areas. James (1971) used results of a principal component analysis to construct an ordination of the distribution of birds with respect to vegetation structure of small trees.

Another approach to ordination arranges plant species abundance in plant communities according to an environmental variable such as percent soil moisture. The resulting species placement along an environmental gradient can be used to infer potential habitat for detailed searches within areas of a community type.

Accurate models based on vegetation, soils, or other habitat information require detailed information about habitat requirements of the plant TES of concern and on the actual occurrence of habitats in the area for which the species' presence is to be predicted. Species requirements may be available if populations have been studied elsewhere. If not, known populations that occur elsewhere may be sampled for habitat information as discussed below under data collection (Subsection IV.C.7). If soils, topography, vegetation, or other habitat information exist or can be obtained at sufficient levels of detail to detect the often narrow habitat requirements of plant TES, then modeling to predict TES distributions for survey or management guidance may be worthwhile.

Highly specialized species, however, may require microhabitats too small to be revealed at the scales of most surveys. For example, narrow strips of riparian soils may not be delineated at the scales of most soils maps. The California bearclaw poppy (Arctomecon californica Torr. & Frem.), a candidate species, is restricted to patches of gypsic soils in the desert southwest (Moore et al. 1993); however, Gypsic inclusions are not distinguished on the general soil map of Mojave County, Arizona where the species is known to occur (Richmond and Richardson 1974). It is worth noting that the qualification is often made in general soil map interpretations that they are ". . . not suitable for detailed planning ... on-site investigations must be made to obtain reliable [soils] information for a specific site or use" (Richmond and Richardson 1974).

If needed habitat information does not exist, a survey to map soils and vegetation and to collect other habitat information will be necessary to obtain information for development of habitat-based models. However, many more habitats would have to be field-sampled than those in which species of interest actually occur. A full vegetation and soils field survey may not be necessary to fulfill the restricted goals of a TES project.

The more straightforward approach outlined below (Subsection IV.B.2) of using qualitative information to locate TES populations in the field, followed by quantitative characterization (during the inventory process) of only those habitats in which they occur, will generally be more efficient than a formal modeling effort.

The focus of the discussions to this point has been on preliminary information that will expedite the location of the TES that occur on the APU. The field survey itself is used to locate and verify which TES occur (or do not occur) and the geographic locations of their populations. Thus, the primary objective of the field survey is to physically locate TES populations and to document their presence.

The field survey is guided by habitat and species information. The information obtained from the previous stages of the preliminary survey (Section IV.A) identifies those species judged to have potential for presence on the installation and the areas to be searched for them. The following stages of assessment focus on locating and documenting, with some degree of confidence, species presence or absence.

The primary objective of the field survey will be to verify, through field searches, the presence or absence of TES judged likely to occur on the APU. Examples of specific objectives related to this primary but general objective are:
1) To locate populations of expected TES, if any.
2) To document localities of TES populations, if found, on appropriate maps.
3) To describe the habitat associated with the occurrence of the TES populations.

Design survey procedures, specify QA/QC criteria, develop a survey budget, allocate funds, specify coordinating personnel, requisition equipment, and recruit qualified field personnel for the survey.

The most common TES survey method is for field personnel to visit known TES populations on the APU or elsewhere to obtain qualitative habitat information. Since the occurrence of the species may not yet be known for the APU, quantitative habitat data collection may not be justified at this point. Once field workers have a "feel" for the habitats in which the TES occurs, they begin field survey in the area in question, guided by available habitat information. For example, if the TES is known to occur in bogs, field workers seek out such areas and search them, using any clues they may have acquired as to which portions or types of bogs are most likely sites. Experienced TES workers develop a "sense" of habitat characteristics of particular TES, which they are able to recognize; however, they may have difficulty communicating useful information to others. Such qualitative methods may in some ways be superior to quantitative methods, at least for the survey stages, because variables for quantitative habitat measurements are often predefined. Important variables or combinations of variables of habitats that may be intuitively discernable by observant field ecologists may not be considered in the design of quantitative data acquisition methods.

Field searches should follow an established search methodology, for example random meanders or transect walks (Nelson 1985) to assure that the range of variation where the species in question may occur is adequately searched. Areas to be searched should be chosen systematically and should cover most, if not all, possible habitat areas. The criteria for search intensity (the level of detail at which potential habitats are examined) if a species is to be considered not present should be specified. In any case, the search strategy must be consistent in that the same pattern is used throughout the search of an area.

NR managers should consider that some annual species will not appear every year, and that some species will be cryptic or unidentifiable except during certain phenological stages and may be missed at the particular time when the habitat is surveyed. Further discussion of survey methods and associated problems are included below in Subsections IV.B.2 and IV.B.5.

In general, the design of the field survey stage will be focused on the TES species and may therefore not include detailed vegetation or floristic surveys; i.e., all plant species will not need to be identified and recorded. Collecting detailed floristic, habitat, and population information at this time is a possibility; however, time and phenological constraints may relegate this work to a separate data collection phase (see Subsections IV.C and IV.D). Design of the survey should therefore specify the level and kinds of plant community and habitat information (e.g., soils or vegetation data) that field crews should collect immediately after locating TES populations. This is a QC function that is discussed further in Subsection IV.D.2.

Minimally, some estimate of the density and areal cover of each TES population should be determined using an established sampling method applied consistently (Bonham 1989). Other desirable preliminary data are estimates of density by life-stage, collection of soil samples, and topographic data (e.g., slope, aspect, and elevation of the site).
Detailed location information should be recorded for each population and marked on topographic maps or aerial photographs. If possible, the exact coordinates of each population's location should be determined using a GPS (geographical positioning system) unit corrected by a base station or with codes that disable selective availability. Field forms or portable data collection devices that standardize the data and location information to be collected should be developed and provided to field crews à prìorì.

Field crews should have explicit instructions on whether or how to modify pre-established search procedures if expected TES are not found. This may involve specification of standards for increasing the intensity of searches in particular areas, when to abandon the search in a given area or season, and when to consult with supervisory personnel.

When population numbers permit, voucher specimens should be collected for the purpose of species verification using standard collection procedures (University of California Herbarium 1975), and should be deposited at an appropriate herbarium for future reference. Before specimens are collected, obtain permits as required by federal or state agencies.
If voucher collection is not possible, detailed descriptions of taxonomic features should be recorded and quality closeup photographs taken. A site visit by a taxonomist should be made to verify the identification. Photographs showing the population in its habitat should also be a part of preliminary documentation.

Study design issues define the intensity of searches and are the result of quality assurance (QA) and quality control (QC) criteria established to ensure consistency, repeatability, and an acceptable level of confidence in the results of the field survey. QA involves detailed attention to the objectives; design; choices of equipment, methods, and personnel; project management; and other factors that affect the outcome and value of a research effort. QC involves the ongoing assessment of data collection and results and their suitability for meaningful statistical analysis as data collection proceeds. A detailed discussion of development of QC criteria for all phases of field research is presented in Subsection IV.D.2. QA/QC determinations should be made prior to any field work.

Predetermination of QA/QC criteria for field surveys assures that methods used will meet the level of confidence at which nonoccurrence of TES will be accepted and documented. How confident can one be that TES populations in fact do not exist on the APU if no populations are found during surveys? This will be a difficult question to answer statistically. While the alpha-levels of 0.01 or 0.05 (for Type I error) are often used for research purposes, it will be difficult to place probabilities on search adequacy given the vagaries of populations that may occur in unexpected areas or at times of the year that are difficult to predict. Thus, survey criteria should not be chosen strictly by research standards. Consideration should be given to the consequences of accepting a hypothesis as false when, in fact, it is true. This type of error (a Type II error) occurs if a field survey finds no evidence of a TES population present when, in fact, there is a population present.

The implications of Type II error will vary according to circumstances. If an area of possible occurrence of a TES is undisturbed, for example, then the question of whether TES actually occur there may be considered of low priority by NR personnel who must focus their energies and funding on avoiding damage to TES populations. Development of adequate QA criteria for field surveys will thus be somewhat subjective, and based on the opinions of experts, the land-use patterns on the APU, and some estimate of the actual probability of occurrence. If a species is expected, but not found, by detailed surveys research could be designed to sample the potential habitat for comparison with habitats of known populations elsewhere. This may then provide a measure of the probability that the habitat on the APU is in fact suitable for the TES.

The budget developed for proposed field research projects should consider all expenses associated with equipment acquisition or leasing, supplies, salaries for temporary and contracted personnel, transportation expenses, insurance costs, housing, and per diem.

Personnel recruitment and training are important aspects of the QA process, because the qualifications of personnel are directly related to the quality and reliability of data that are obtained. Coordinating personnel (personnel who will coordinate and supervise the survey effort) should be permanent NR staff members or contractors with training and experience in field research design and supervision of vegetation studies. Field survey personnel should have field experience and familiarity with the local flora, particularly with the species in question or with other TES. Other skills should include map and aerial photograph interpretation, work with data collection devices (e.g., data loggers, GPS units) and experience with field sampling methods for TES. It should be noted that field personnel may locate unexpected TES species or may locate one species while searching for another. This provides a good reason for hiring taxonomically skilled individuals who have knowledge of local species.

Often field personnel are hired on a seasonal or temporary basis to reduce costs. However, seasonal hiring often does not assure continuity of personnel from one field season to the next. Additionally, individuals available for seasonal work are often inexperienced. Contracting with an organization that can provide qualified field workers, such as private consulting firms, the Nature Conservancy, or a state Natural Heritage Program, is a hiring option for field work that must be conducted seasonally.

In addition to selection of methods and personnel, selection of equipment is a part of the survey design and QA phases. Equipment acquisition should be addressed prior to field work. Equipment includes all materials necessary for completion of the field work (e.g., tapes, quadrats, soil containers, maps, aerial photographs, compasses, GPS units, clinometers, cameras, portable computers, field forms, plant presses, and microscopes). Also to be considered are camping equipment, suitable vehicles, housing, and equipment needed for data analysis (computers, software, plotters, office space). Equipment acquisition can represent a sizable portion of the project budget. Training sessions must be planned and conducted to familiarize personnel with the equipment and methods selected for the survey.

In some situations it may not be feasible to conduct field surveys. This may be the case where sufficient funds are not available; for military installations, where contaminants, live ordinance, or training activities make field work impractical or dangerous. Under such scenarios, work must be delayed until field surveys can be adequately and safely conducted.

Occasionally, predictive habitat-based modeling of possible TES occurrence may be considered as an alternative to field survey (see Subsection IV.A.3). Habitat-based models are seldom highly predictive because species occurrence depends on factors other than presence of suitable habitat. Thus, eliminating the field survey bypasses the crucial step of verification of model predictions. The NR manager who uses models as the sole basis for land-use decisions must accept the possibility that the result may be protection of areas where TES do not occur, or inappropriate land uses where TES do occur. Use of predictive models to make decisions affecting TES may be questionable with respect to the legal requirements of the Endangered Species Act.

Conduct the field searches at appropriate seasons and in appropriate habitat. Document TES occurrences.

As mentioned above, some species will be visible or identifiable only at certain seasons of the year. Most plant species will be most easily found when in flower, and some may be identifiable only when in fruit (e.g., species of Brassicaceae, Apiaceae, and Fabaceae). This makes the timing of surveys crucial, and underscores the need for a thorough knowledge of the TES prior to field work. Even with these precautions, however, phenologies of species may be difficult to predict from year to year, and surveys may need to be repeated several times during a season to locate populations at an identifiable life-stage. In addition, some species may not flower and fruit every year [Desert annuals, in particular, may not appear each season, and ascertaining their presence or absence can be a complex undertaking]. It may be impractical to conduct surveys in dry years when annual TES are not expected to germinate. This is another situation where consultation with local experts may facilitate surveys, as they may have knowledge of conditions that are favorable to the species. Reference to the phenological condition of nearby, known populations, if any, should help to guide field searches for such species.

Geographic, taxonomic, and other information for documentation of TES presence, as well as specification of preliminary population and habitat data to be obtained by field survey crews, were discussed in Subsection IV.B.2. Documentation of TES absence will involve geographic descriptions of the areas searched, description of the intensity of searches, QA/QC criteria followed, and methods used, as determined in the field survey design phases.

For NEPA (National Environmental Policy Act, 1969, 14U.S.C. 4321-4347) compliance (e.g., for inclusion in an environmental impact statement (EIS), the following format to report the results of field surveys is recommended:

1) Project description including detailed maps.
2) A written description of the biological setting, referencing the plant community nomenclature used and a vegetation map.
3) Description of the survey methodology.4)Dates of field work.
5) Results of the field survey with detailed maps.
6) Assessment of potential impacts.
7) Discussion of the importance of the populations, with consideration of nearby populations and the distribution of the species
8) Preliminary recommendations for mitigation and impact reduction.
9) A list of all species identified.
10) Copies of field forms.
11) Names of the field workers.
12) Lists of references, persons contacted, herbaria visited, and the location of voucher specimens.

The term "inventory" is often used interchangeably, and in conjunction with, "survey." This creates some degree of confusion as to distinctions between these two terms. Some definitions, for example those used in Bureau of Land Management Manuals 1734 and 6600, consider inventory to be a "periodic" activity, thereby further confusing "inventory" with "monitoring." In this Protocol, the term "inventory" refers to a one-point-in-time assessment. As such, it is distinguished from monitoring (see Subsection IV.D), which uses the results of quantitative data collection repeated through time to detect changes in population parameters, associated vegetation, habitat, or environment (Palmer 1987). Inventory, like survey, may or may not be considered a separate activity from field data collection, although inventory is commonly thought of as a field data collection activity.

The TES inventory, then, may include listings of the TES present on the APU geographic information for TES populations, counts of individuals within populations, or more extensive quantitative data collection on individuals, populations, and habitats that can be used as a basis for monitoring. Thus, in this Protocol, the purpose of an inventory is to describe the important characteristics of species, populations, and their habitats within the local setting of an APU. For simplicity in presentation, the inventory is presented first as a compilation of the results of the preliminary and field surveys, and second, as encompassing quantitative data collection.

Primary objectives of the inventory will be to compile a listing and to determine the status of TES populations present on the APU. More detailed objectives, to be defined by the NR manager, will determine the kinds of data to be collected in the inventory stage.
Objectives of the inventory might be to:
1) Provide baseline data on population status, abundance, and distribution.
2) Characterize habitat where populations are found.
3) Map locations of TES populations.
4) Catalog TES present.
Often, data collected during the inventory stage will be used as an initialization of the monitoring effort.

Compile an inventory of TES documented to occur on the APU. Thoroughly document searches for expected species that are not found.

The inventory of TES located on the installation, coupled with geographic information and baseline data, will be the basis for future management decisions that may affect land use activities and will guide any further research. While additional research may be required to determine the actual effects of human-caused activities on TES populations, the inventory also identifies where current or projected human activities and locations of TES populations coincide. Thus, it provides a framework for further study and for prioritization of funding allocations.

Species may not be found during field surveys, though their occurrence is expected because suitable habitat occurs on the APU. Field survey methods, including search intensity and locations searched, should be thoroughly documented in such cases. Field methods should be reviewed by U.S. Fish and Wildlife Service (USFWS) personnel prior to land-use decisions to ensure that survey procedures comply with their regulations pursuant to ESA.

C.3. Populations of Special Concern and Preliminary Management Recommendations

Identify populations of special or immediate concern because of observed or qualitative indications of low densities, poor reproduction, or poor survival. Compare areas and habitats of concern with current and planned land uses.

Presence of TES on an APU will usually indicate that some level of management response is needed, particularly if human activities coincide with the locations of populations. Analyses may include exploratory analysis of preliminary habitat and population data in relation to land-use patterns. Some form of spatial overlay (e.g., GIS or spatial statistics) analysis to detect areas where TES populations and current or projected management or land use activities coincide will be of particular interest. Also of value would be information on TES populations that occur in areas of past land use activities, as this may indicate favorable responses to particular kinds of disturbance.

In consultation with species experts, make interim recommendations regarding possible land-use conflicts between land use activities and the security of documented TES populations.

Results obtained from the survey, other than presence or absence, should be regarded as tentative with respect to the effects of land uses on apparent trends in population numbers. Both causation and population trends may not be evident without detailed studies over time. Accurate specification of the real problem may thus require further study. Conclusions should not be made before adequate and appropriate data exist. Consultation with species experts may help to determine negative effects discernable in data collected during surveys, and thus provide a framework for preliminary management recommendations to temporarily modify existing land-use.

An example (personal communication from Dr. Wayne Leininger, Colorado State University, Fort Collins) illustrates the relationship between early qualitative impressions, monitoring over time, and management response. The threatened orchid, Spiranthes diluvialis, was initially managed by exclusion of domestic livestock grazing within its habitat. However, monitoring revealed subsequent population declines, and grazing disturbance was concluded to be necessary for maintenance of appropriate habitat conditions for this species. Military training exercises are also known to benefit some plant TES (Shultz and Shaw 1992).

Prioritize species concerns for allocation of research funds and personnel among TES occurring on the APU.

Some TES, for example species endemic to only an APU, species that have a major portion of their extant populations located on an APU, or species that occur over large areas of an APU, will be of higher concern than others. Populations that appear to be robust and that are not at risk from land use activities may be viewed as low in priority. In some cases, specific questions that relate to the status of populations or their distribution or habitat requirements may have been raised during the survey phase that need additional study. Prioritization of research efforts facilitates allocation of funds and other resources where they are most needed.

A draft U.S. Army Construction Engineering Research Laboratory TES working group document called for the development of an investment strategy that specifies criteria for the prioritization of research fund allocation for research and conservation efforts among species and among military installations (Threatened and Endangered Species Working Group 1995). Several factors are considered in the document, including the vulnerability of a TES to military activities, the degree to which an installation can accommodate research efforts, and the priorities accorded the TES within other agencies. Research and management activities that are being conducted by other agencies or installations for particular TES can be used to prioritize fund allocations, since such activities may represent opportunities for combined regional efforts within the context of an ecosystem management approach.

Design population and habitat data collection procedures, specify QA/QC criteria, develop a budget, specify coordinating personnel, allocate funds, requisition equipment, and recruit qualified field personnel for inventory data collection.

As mentioned above, habitat and population data (treated here as part of the inventory process) often may be efficiently collected concurrently with field survey procedures if time permits. Often, however, the window of opportunity during which populations can be surveyed, located, and documented may be brief. This is also true for population data (e.g., reproduction as indicated by flower production or seedling establishment) where individuals must be counted and categorized by life-stage. The decision of whether survey and data collection (inventory) are conducted simultaneously or separately should rest with coordinating personnel. Decisions will usually be made on a species-by-species basis to achieve maximal efficiency.

Data collection procedures and methods established at this time or at the field survey stage may also serve as the first in a series of periodic monitoring measurements. Methods and procedures selected should, in such cases, be designed with such longer term data requirements in mind. Designs should therefore consider permanent plot markers or grids, tagging individuals of perennial species, and other means that provide for repeated measurements over intervals of time. Data collection and sampling design may be improved by use of designs that were previously used in similar vegetation types in the region. Refer to Subsection IV.D.2 for discussion of strategies for the development of data collection procedures and QA/QC criteria for monitoring studies.

The intensity of data collection and monitoring for plant TES should also be selected on an individual-species basis, with consideration for the security of populations and the priorities accorded each TES by NR management personnel. For example, species of high priority because of immediate threats to populations from land use activities or because of critically low or declining populations may require intensive study to isolate causes and effects. Stable or increasing populations may be monitored infrequently to provide a warning of any decline abundance.

Standardization of sampling and monitoring methodologies and intensities across all TES present on an APU may be wasteful by providing unnecessary information for secure populations and too little information on species that require intense study. Such decisions should be made carefully, however, and in view of the quality of data that are presently available. As discussed below under monitoring, the size of populations as determined in the inventory stages may provide some indication of their status and trend, but only monitoring of populations will reveal actual change. Species priorities and methods should be reviewed on a regular schedule as new information becomes available through monitoring, as part of the feedback loop of adaptive management.

Design of inventory data collection procedures will include specification of the kinds of data to be collected and consideration of QC criteria to be used in evaluating the progress of activities.

QA/QC criteria for all monitoring phases are discussed at length in Subsection IV.D.2. The general logistics of budget, hiring, and equipment acquisition are comparable to those of the field survey phase (Subsection IV.B.4). Data collected in the inventory and survey phases will often be the first in a series of measurements over the course of a monitoring project; however, in some cases baseline data may be desired for species whose priority does not warrant monitoring. In such cases, study design and QA/QC for survey, inventory, and monitoring will still need to be integrated.

Habitat data collection requires technicians who have the taxonomic skills to identify plants present in sampled areas, and the sampling skills needed to provide accurate data. Often, workers for data collection are hired on a seasonal basis because of time constraints on permanent NR staff members (see Subsection IV.B.4). The actual characterization of TES habitat is a data analysis step that should be performed by permanent personnel or contractors with skill and experience in statistical methods.

Collect data on relevant habitat, environmental, and vegetation characteristics for all or a representative sample of documented populations of each TES.

Habitat information relevant to plants includes soil texture, depth, and nutrient characteristics, topographic information (elevation, aspect, slope), animal use, ground cover, disturbance levels, moisture regime, floristic and structural composition of the plant community and the seral stage of the potential natural community. Habitat information may be collected at several levels of intensity, as selected by coordinating personnel. For example, vegetation may be characterized by dominant species only (e.g., oak-hickory forest) if this level of information is viewed as adequate. Concurrently, data that allows calculation of frequency, density, percentage cover, and species height strata for the TES and all or some of the associated species may be needed. Data collection should follow an established sampling design (see Bonham 1989), should be designed to suit the analysis methods to be used, and should meet quality assurance criteria for confidence in results.

Habitat data are useful to provide basic information about species requirements, to predict potential locations of additional TES populations, to predict distributions using spatial statistics or other modeling techniques, to locate potential sites for reintroduction, and to provide baseline information for restoration efforts and propagation (see Subsection IV.E.2).

Collect population data by size class or life stage, at selected levels of detail, on all or a representative sample of populations and during appropriate seasons.

Population data are collected to enable monitoring of the status and trend of populations over time and to project change in population growth using population modeling techniques. Population data for monitoring may be collected at several levels of detail, including a) simple census of the density of individuals over time, which can provide an estimate of population stability from changes in population size; b) census of reproductive individuals over time, which can project population trend; and c) full demographic monitoring of the recruitment (birth) and death rates within populations, which allows use of population matrix models to project population trend and to identify life stages that most affect the growth rate of the population.

Data collection procedures (see Subsections IV.B.2, IV.C.7, IV.D.2) should be designed with consideration of the kinds of analysis to be conducted to avoid wasteful collection of unnecessary information. Data analysis for TES studies will fall into several categories: a) exploratory data analyses, b) habitat characterization and trend, c) population status and trend, d) distribution and spatial analyses, and e) analyses of additional research data.

Exploratory data analysis involves the use of statistics and graphical displays of data points to explore patterns that may emerge in a data set (Tukey 1977, Ellison 1993). For example, plotted frequency distributions of life stages (e.g., seedling, reproductive, post-reproductive) measured in a TES population may permit initial estimates of apparent trend to be later verified by monitoring. Scatterplots and ordination diagrams (see Gausch 1982) may be of value in exploring bivariate and multivariate data sets.

Habitat characterization is the identification of the components of habitat important to a species. Multiple regression (ter Braak and Lawman 1987), correlation analysis, multivariate techniques such as ordination by principal components analysis or detrended correspondence analysis (ter Braak 1987, Ludwig and Reynolds 1988), and permutation methods (ter Braak 1987) can be used to identify consistent relationships between a species and habitat factors that affect its distribution.

Population demographic data, such as reproduction and survival rates, can be used to develop stage-based life tables and, using stage-based matrix simulation models (Lefkovitch 1965), to estimate population trajectories and the life stages most affecting population change (Menges 1986, Schemske et al. 1994). As described above under population data collection (Subsection IV.C.7), several possible levels of population data collection lead to different kinds of analysis. The most basic analysis of population density data collected by life stage provides an estimate of the current status of populations (Schemske et al. 1994).

Spatial statistics (Burrough 1987, Turner et al. 1991, Cressie 1993) can be used to describe and analyze spatial variation in population and habitat parameters. Such analyses can include graphical displays of the geographic locations of populations and correlation of distribution data with habitat and population data. GIS techniques, as discussed in the section on prediction for field survey (Subsection IV.A.3), are useful for overlay analyses that may help to identify consistent correlations among topographic features and species distributions.

The analysis techniques discussed to this point are primarily descriptive (Eberhardt and Thomas 1991), or they are predictive in the case of population modeling. Comparative and experimental data (i.e., from comparisons of spatially or temporally separated measurements, or from manipulative, replicated experiments) enable the use of inferential statistics for discerning differences among measurements or treatments. Analysis of variance (ANOVA) and Analysis of covariance (ANCOVA) can be used to separate the effects of various levels of environmental variables (a comparison of responses of TES in undisturbed vs. disturbed habitat) (Potvin 1993). See Subsection IV.D.5 for references on analysis of temporally separated measurements.

Data analysis should be performed by NR staff members or specialized contractors skilled in statistical methods. Every effort should be made to assure continuity between design and analysis of field studies; data collected should be appropriate for the kind of analysis that is planned and should be intended to answer specific questions as previously determined in the identification of objectives for the project.

Report on field work, data analysis, results, predictions and conclusions.

Publication of the results from an inventory for TES expands the knowledge base about target species, whether found or not. When possible, peer-reviewed outlets should be selected. This informs NR managers in different agencies and services, disseminates information to the scientific community, and provides a mechanism for peer critique and feedback. At minimum, internal reports should be made available to research libraries with the sponsoring agency's approval. See Subsection IV.D.6 for a more detailed discussion of formal research presentation.

At this stage, preliminary recommendations should be made as to possible modifications of land-uses to eliminate or avoid impacts to TES and their habitats. Definitive conclusions as to the effects of land use activities or NR management strategies on TES populations will not be possible prior to monitoring, although in many cases effects may seem obvious (see Subsection IV.C.3). Some species respond positively to disturbance activities and may therefore decline if the disturbance is stopped or substantially altered. Nevertheless, the first response of management to a newly discovered TES population is often protection from all disturbances. For this reason, coordinating NR resource personnel should report results at this stage to relevant land managers on a tentative basis, and should consult with species experts to assess likely response of populations to disturbance.

Monitoring is the systematic collection of data over time to detect changes in relevant population or habitat attributes, to determine the direction of those changes, and to measure their magnitude.

Monitoring involves re-measurement of selected variables at determined time intervals. Measured characteristics of TES should be selected that will most likely respond to change in one or more habitat variables. Monitoring efforts assume that a defined and measurable TES characteristic exists as a baseline against which change can be measured.

Most studies of TES utilize repeated measurements during monitoring as the main source of field data. This is necessary because TES populations are normally too rare to permit controlled experimental manipulations in natural settings. The most reliable information on the status and trend of TES populations and habitats will be provided as monitoring through time proceeds. Preliminary estimates of population or habitat status obtained from initial data collection should be revised, if necessary, as results of monitoring studies become available (see Subsections IV.B and IV.C). For example, monitoring could reveal that a small, apparently precarious TES population actually remained stable over time, or that a larger, apparently robust population showed declines in density or recruitment.

The objectives to be achieved in the monitoring of TES populations and habitat will relate to the detection of change in parameters over time. Specific objectives might include:
1) To determine trend in population numbers.
2) To detect a specified change in numbers that will indicate a demographically secure population.
3) To provide data for modeling population trend.
4) To identify effects of terminating habitat disturbance on TES population size.

Design and schedule population and habitat monitoring procedures, specify QA/QC criteria, specify coordinating personnel, develop a budget, allocate funds, obtain equipment, and recruit qualified field personnel for data collection.

The design of a monitoring study is based on the objectives previously established. Types of data to be collected will answer the questions proposed. For example, monitoring of population status obviously requires data on its demographics (see Subsections IV.C.5 and IV.C.7). Variables that can be monitored in TES studies include change in the composition of plant communities associated with a TES (e.g., succession), which requires data on vegetation such as cover; availability of suitable habitats, which requires initial data on habitat requirements of species and periodic data on the status of available habitat; the invasion or increase of exotic species, which involves measurement of the floristic composition of the community; changes in moisture regimes, requiring weather data and periodic measurements of soil water; and changes in population growth rates, requiring demographic data for the population in question. Detection of population change will be of particular relevance for TES management efforts.

Detection of change in these variables requires attention to issues that may be decided at a stage prior to formulating a "monitoring project" as such. The appropriateness of the kinds and levels of data collected at the earlier field survey and inventory stages, the ability of chosen analysis methods to detect change, and the biological significance of the parameters measured to TES populations, should all be considered with the knowledge that ultimate establishment of a monitoring program may occur. In short, although survey, inventory, and monitoring are treated in separate sections of the Protocol to emphasize the unique functions of each stage, NR managers involved in TES work should keep in mind that the design of data collection procedures at the stages of survey and inventory may represent the first in a series of measurements for a long-term monitoring project. Since most useful TES research will involve monitoring, this implies that design of all phases of the TES project should be considered as a whole. Thus, issues addressed below that pertain to, for example, QC and the sensitivity of methods to detect change relate to the data collection phases of the survey and inventory stages as well as to the monitoring design itself.

The schedule of monitoring (e.g., yearly, every 5 years, etc.) should be designed, like any aspect of TES research, with consideration of the phenology and natural history of the species in question, and with reference to relevant literature and the opinions of experts.

Quality assurance/quality control criteria should be established that can provide reliable data at confidence levels sufficient to detect ecologically important changes in populations and habitats, but with acceptable Type I and Type II error levels that meet other requirements of the monitoring project, including cost criteria (see Subsection IV.B.4, QA).

Individuals with specialized knowledge of species taxonomy and ecology, habitat, TES legislation, and study design and analysis, may be required for the successful design and implementation of a TES monitoring program. Such expertise may be beyond the experience and training of NR managers. For this reason, two preliminary stages preceding the finalization of a monitoring methodology are recommended. These are: a) Conduct a meeting of coordinating NR managers, other local experts, contractors, and professionals in relevant fields such as ecology, soil science, and hydrology. This group would meet in a workshop forum to address the details of inventory and monitoring designs and appropriate QA/QC criteria; and b) conduct preliminary studies to test proposed methods for sensitivity in detecting change, statistical confidence levels obtainable from data collected, and other issues such as time- or cost-benefit, ease of use, and training and implementation problems associated with selected methods.

The workshop is intended to define the specific TES problem(s) of concern to NR personnel, to determine specific characteristics of the problem(s), to list the kinds and amount of TES population and habitat data needed for decision-making, to set levels for detection of changes in TES populations and habitat and to decide upon desired and meaningful levels of precision, accuracy, and confidence levels for each set of data to be acquired. Changes can be described from ecological and management perspectives and will depend on the critical components of the system that may be affected by land use or natural processes through time. The outcome of the workshop exercise provides information for the development of a QA/QC Plan to be followed in monitoring changes of TES populations and associated habitat characteristics.

Exact specification for each component of monitoring is incorporated into a QA/QC criteria format to assess and control the value and utility of population and habitat data collected. Specific field data sets to be collected and analyzed will be derived from the output of the workshop.

Information needs of the NR staff will dictate specific uses and requirements of the monitoring program. Natural resource personnel must provide guidance during the workshop for specification of the effects to be detected in monitoring. For example, key associated plant species and ecologically important species (if different) might be identified at the Workshop. The ecological importance of defined habitat and vegetation characteristics, and their relationships to the TES in question as identified by workshop participants, should then be determined and translated into acceptable statistical significance, with use of measurement methods that provide unbiased or minimally biased estimates of means and variances for the characteristics measured. The goal of the workshop should be consensus on measurements needed and how measurements are to be taken.

NR personnel or contractors should evaluate the workshop report to verify appropriateness of suggested measurements. Literature searches may provide sufficient information to make professional decisions to reduce bias. A small preliminary field study in a portion of a TES population to evaluate methods suggested in the workshop can be used to evaluate workshop recommendations. During the workshop evaluation, those involved must keep in mind that data collection for the larger study should follow an experimental design that yields statistically valid results for the vegetation-environmental characteristic(s) measured, including spatial distribution effects.

Before field measurements are obtained by the final methods selected, NR personnel or contractors should assess important aspects of the methodology, e.g., comparing random vs systematic sampling methods of plots, lines, or whatever equipment is to be used? Will spatial statistical procedures be used to describe population distribution characteristics and associated environmental and habitat factors? Will the answers to these questions affect QA/QC requirements as determined from the workshop? The final field sampling design depends on the answers to these kinds of questions.

A second workshop to discuss results of the field data after analyses have been completed with the same group may be needed. Discussions would include additional ecological and statistical interpretations. For example, concerns may arise with respect to meeting QA/QC criteria. Decisions can then be made about the final monitoring design to include selection of permanent markers, photo-points, equipment to be used, TES population and vegetation-environment characteristics to be measured, and time intervals to be used in data collection. Discussion may include whether or not a need exists for the incorporation of ecological risk into the studies. Such a discussion should be structured by considering the ecological consequences of decisions made based upon information obtained in the monitoring process.

Additional analysis of the preliminary baseline data obtained for monitoring purposes (e.g., from data collection at the survey or inventory stages) may be conducted to determine the power of the sampling design for detecting change in TES populations. This exercise will provide useful discussion prior to finalization of monitoring plans. Selection can then be made for the final design characteristics of the monitoring plan.

Development of a detailed methodological protocol to be followed for data collection, analyses, interpretation, and reporting will be a desired outcome of the synthesis workshop. Details needed in the specific monitoring protocol will emerge as decisions are made by the NR staff about TES of concern, ecological interpretation of data obtained from monitoring, and use of such data in management decisions (e.g., restoration of habitat, etc.).

Efficient monitoring methods for TES populations or habitat are characterized by: 1) stability, 2) power, and 3) robustness. Stable methods are those that will not result in false conclusions of change when no change has occurred. Powerful methods detect change when it has occurred. Lastly, methods are said to be robust when measurement techniques provide data that are independent of the technique used; e.g., plot size or transect length (Brady et al. 1993b). If observer biases are observed when using a sampling technique, additional training and calibration for observers will be needed.

Some monitoring methods result in data containing various sources of error that can lead to improper management decisions. This is the case when data falsely indicates that a change has occurred (in statistical terms, Type I error) or when an actual change is undetected (Type II error) (Tanke 1984, Tanke and Bonham 1985). The consequences of either kind of error can be serious when assessing dynamics of TES populations. Therefore, data collection should be designed with stringent QA/QC and clear specification of confidence levels expected in the data analysis stages. A properly designed monitoring program will minimize risks associated with these kinds of error.

Small sample sizes often are a limitation in TES studies. Be aware that the power of a monitoring design for detecting change is a function of several factors, including variability inherent in the data, experimental design, sampling design, and selected levels of Type I and statistical errors. The level of Type I error selected determines the level of Type II statistical error (failure to detect change when it has occurred) that is associated with the given experimental or sampling design. The use of power curves to evaluate methodology for detecting changes or differences in variables has been discussed by several authors (Tanke and
Bonham 1985, Gerrodette 1987, Green 1979a, Lipsey 1990, Brady at al 1995, Rice et al.1995). Because of the consequences of making decisions based on Type II errors, Tanke and Bonham (1985) provided mitigating methods of selecting acceptable Types I and II errors. Brady et al. (1990, 1991, 1993, 1995) used Type II errors and computer simulation procedures to derive power curves for evaluation of vegetation measurement techniques.

Monitoring designs should be developed to balance financial resources simultaneously with data requirements. One alternative is to stipulate required levels of stability and power, and then meet these requirements at least cost. The other alternative is to stipulate available financial resources and then maximize stability and power of the design subject to meet these constraints. The latter requires that Type I and Type II error values be flexible rather than pre-selected. Such alternatives should be discussed during the first workshop.

The logistical details of specifying personnel, budgeting, acquiring equipment, and hiring and training personnel are fundamentally the same for monitoring as those for the field survey and inventory stages, (see Subsections IV.B.4 and IV.C.6). However, these details are of particular concern in monitoring because the effort will extend over a period of several to many years, and in some cases could continue indefinitely or until species are delisted. Allocation of off-season time for data analysis and interpretation, or arrangements for contractors to perform this work, is necessary. A monitoring effort will typically encompass several fiscal periods, complicating the management of funding for continuation of the project.

Initiate and conduct the monitoring and analysis schedule for known populations to determine the current and projected status and trend of TES populations.

The status and trend of populations has been argued to be the most important and significant of data variables that can be generated in the initial stages of a TES study (Palmer 1987, Schemske et al. 1994). This is because the fundamental concern of ESA is the security of TES populations. Data collection and analysis methods to estimate population status and to project population trends rely upon direct measures of demographic parameters. Such information is crucial for determining appropriate management response.

Projections of habitat decline, on the other hand, infer responses of populations via correlated habitat factors. While habitat availability is obviously a major factor in the security of TES populations, the most direct measure of population trend is provided by demographic data. Habitat monitoring, however, may provide an "early warning" in cases where demographic data are not available. Such uses assume that measured change in characteristics of the habitat can be directly related to population change. Data collection and analysis for determination of population status and trend were discussed in Subsections IV.C.7 and IV.C.8. See also Owen and Rosentreter (1992) for a discussion of demographic data collection for monitoring.

Analyze monitoring results periodically for changes in populations possibly caused by land uses. Continue some level of monitoring of populations and habitat as a feedback mechanism for timely adaptive management in response to changes in population status and trend.

The value of monitoring in a management context is that detected change in TES populations or habitat can provide feedback for timely species management or changes in land use. Monitoring also allows agencies to direct further research to answer specific hypotheses about the possible causes of population changes. Monitoring results should be examined regularly as new results are produced. Regular analysis of monitoring results capitalizes on the utility of monitoring as an "early-warning" system for management.

Repeated measures analysis is an appropriate technique for determination of how population parameters change in different populations of a species in different habitats over time (von Ende 1993). Stage-based population matrix modeling, as suggested in Subsection IV.C.8, can isolate the life-stages of a TES population that most affect population trend, thereby providing direction for management, further research, and mitigation efforts. References on population modeling are provided in Subsection IV.C.8.

Report on and publish results of periodic monitoring efforts and make management recommendations.

As with any research effort, reporting and publishing results provides for the dissemination of valuable information obtained from TES studies. Reports and scientific papers should include methodologies, analysis, and interpretation of the results and should be presented to scientific peers or department personnel for criticism, suggestions, and alternative interpretations before publication.

Submissions for formal publications should follow the format specified by the selected journal or agency, although generalized formats are discussed in Ambrose and Ambrose
(1987) and Day (1988). A typical internal report format that includes management recommendations follows:

1). Introduction
2). Methods
a). Study area
b). Methodology
3). Results and Discussion of Results
4). Summary and Conclusions
5). Management recommendations
a). Status of TES
b). Protection needs
c). Additional monitoring and research needs
6). References
7). Appendices

Continue monitoring over the long term, possibly on a revised schedule based on emerging information. Design and implement further research, if necessary, to determine the causes of changes in both extant TES plant populations and in potential habitat.

As data emerge from monitoring, the intensity and frequency of monitoring may need to be revised. For example, species priority may be changed or additional research may be needed to enhance species management efforts. Examples of additional research might be physiological, ecological, genetic, or landscape studies designed to answer specific questions about TES populations. Examples of issues that may require further study include species responses to specific land uses or contaminants, the long term viability of populations (Menges 1990), the status of populations that occur on nearby lands administered by other agencies, the effects of inbreeding in small and isolated populations, the effects of adjacent populations of non-native or native congenerics that may result in hybridization (Ellstrand 1992), and other matters related to the security of TES populations. While experimental field study may be problematic for TES populations, greenhouse experiments of germination or other characteristics may be practical.

Landscape-scale modeling, such as models of succession or disturbance may be of value to predict future habitat conditions for TES populations. Such approaches may be useful for species that require particular seral stages for survival, especially in areas where natural disturbance regimes have been altered (e.g., by fire suppression or elimination of grazing). Populations that require open habitats, for example, will decline if succession proceeds toward a more closed vegetation overstory. If adequate habitat in suitable seral stages does not become available elsewhere, populations may decline. In such cases, models of vegetation change over time may provide an estimate of future availability of suitable habitat. Such models may indicate a need for artificial disturbance to generate necessary seral stages.

For cases where populations or habitat of TES of particular concern have declined, consider manipulative intervention techniques.

A primary goal of TES study and management is to anticipate population trend to avoid costly and possibly ineffective intervention measures, such as habitat restoration, captive propagation, artificial pollination, artificial disturbance, and population re-introductions. These techniques, however, are available in cases where population declines must be arrested or habitat loss mitigated for compliance with ESA or NEPA. Specialized expertise will usually be required, both in the planning and implementation stages. Long-term species survival can often be enhanced in an area by artificially increasing the number of populations present, thus spreading the risk of losses due to stochastic natural events or accidental human caused disturbance.

Individual TES populations may represent subunits of a group of populations, or metapopulation, that are linked genetically by immigration and emigration. Treating the metapopulation as the unit of concern may clarify species management activities in some cases. A decline in an individual population may not always call for heroic measures to halt the decline if the metapopulation is faring well. This may be the case, for example, for species dependent on successional stages of vegetation, where populations disappear as succession proceeds but new populations arise in other areas that are in appropriate seral stages. These issues, however, have ethical and legal implications requiring coordination with other agencies, especially if subpopulations occur in different jurisdictions. Any considerations will require a true ecosystem management approach at landscape scales.

The point at which an agency's obligations under ESA are complete will be difficult to define. This Protocol treats the monitoring and management of TES as an ongoing, cyclic process of feedback among species management, land-use management, and emerging information from population and habitat monitoring (Fig. 1). This process is known as adaptive management.

So long as a species is listed as threatened or endangered by USFWS, it is protected under ESA against harm, and against harm to its habitat, that result from human activities. These protections may be far-reaching. For example, construction may alter watershed hydrology that in turn, affects moisture regimes and possibly critical TES habitat. Human use of the lands in question implies a need for ongoing efforts to monitor the status of TES populations and their habitat, so that managers are alerted to TES declines.

The ultimate goal for listing of a species under ESA is to promote recovery of the TES. When a species has recovered and it is no longer viewed as threatened or endangered, it is delisted by USFWS. The ESA does not specify quantitative criteria to determine when a species can be considered recovered. However, recovery plans approved by USFWS outline specific measures and research to be performed towards stabilizing or improving the status of a species. The recovery of a species is generally an issue of coordination of efforts on more than one population across several land ownerships, and is therefore best addressed by a multi-agency recovery team, in the context of a holistic ecosystem management approach to species recovery, rather than by natural resource managers whose concerns are the individual populations in their charge.

The matter of prioritization of TES on an APU is also related to the level of concern for individual species. In some cases, an agency or department may want to take primary responsibility for TES species that have a major portion of the species range on their land, or that are endemic to lands administered by that agency or department. Such efforts should be coordinated with USFWS and the recovery team (if one has been formed for the TES in question). Most questions of TES security, prioritization of species concerns, designs for additional research, and the need for monitoring of species that are apparently not impacted by management activities or by other land uses should be approached in consultation with members of recovery teams and with USFWS.

Part II: Models for Inventory, Monitoring, and Management
of Threatened and Endangered Plant Species
Sections 1-IV
Sections V & VI

Document placed on the Web Monday, April 2, 2001
This page has been accessed times.