4 Goals and Objectives Now and Into the Future

Societal values are often the primary factors influencing the goals and objectives of a monitoring or management plan (Elzinga et al. 2001, Yoccoz et al. 2001). Consequently, it is important to understand what goals society has for the resources involved. There are many guidelines available that document how best to identify, engage and understand stakeholders in an issue, empower them in decision-making as the plan is developed, and integrate them as key partners in the adaptive management process. Yet these are far from simple tasks, and even if societal values are fully understood and integrated, these values change, sometimes abruptly. Societies, cultures and the expectations of their members evolve as surely as do species and ecological communities. This presents a daunting challenge for those charged with developing a monitoring plan, because the selection of the species and habitat elements, and the scales over which they are measured, must be selected now in the absence of knowing if these will be the correct parameters to have measured 5, 10, 20, or 100 years from now (Figure 4.1).

In light of this, program managers should work with stakeholders to identify easily understood indicators of state variables (e.g., populations) and their state systems (e.g., habitat) to increase the odds that they will inform the decisions of future managers in a meaningful way. This can be difficult as there are often many options to choose from: Whitman and Hagan (2003) developed a matrix of 137 indicator groups by 36 evaluation criteria as a means of indexing biodiversity responses to forest management actions. A good rule of thumb for ensuring that yours are easily understood is to keep in mind that as indicators begin to span multiple species, multiple times, and multiple areas, clearly articulating goals, objectives, and uses for a program becomes increasingly complex. Although the numerous obstacles and complicated decisions are daunting, making the effort to understand and incorporate societal values is often the only way to develop indicators of change that will be meaningful to society and meet specific monitoring goals and objectives.

Targeted Versus Surveillance Monitoring

Before the process of setting goals and objectives begins, one should have a clear idea of what monitoring is and is not. In their review of monitoring for conservation, Nichols and Williams (2006) argue that monitoring should be equivalent to any scientific endeavor; complete with clearly defined hypotheses that should be produced through deductive logic and be postulated well before any data are collected. They go on to discuss what monitoring is by contrasting two distinct approaches: targeted monitoring vs. surveillance/“omnibus” monitoring (Nichols and Williams 2006). Targeted monitoring requires that the monitoring design and implementation be based on a priori hypotheses and conceptual models of the system of interest. In contrast, they suggest that surveillance monitoring lacks hypotheses, models or sound objectives (Nichols and Williams 2006).

Surveillance monitoring, however, is the more common of the two and often involves data collection with little guidance from management-based hypotheses. In many cases, these types of programs focus on a large number species and locations under the assumption that any knowledge gained about a system is useful knowledge. Surveillance monitoring has been criticized as “intellectual displacement behavior” because it lacks management-oriented hypotheses and clearly defined objectives (Nichols 2000). The primary, often unstated, goal of most surveillance programs is the continuation of past monitoring efforts and the identification of general population trends. Once a trend is detected, usually a decline, management options such as immediate conservation action or undertaking research to identify the cause of these declines are generally implemented (Nichols and Williams 2006). The main limitations of this approach are a dependence on statistical hypothesis testing for initiating management actions (i.e., an insignificant trend would lead to no management), time lags between an environmental change and a population response, costs and resource availability, and a lack of information on the causes of decline (Nichols and Williams 2006).

Despite its limitations, however, surveillance monitoring should not be viewed as a wasted effort. Many of the large-scale monitoring programs covering large geographic regions and estimating changes in numerous species or communities, such as those discussed in Chapter 2, could be considered forms of surveillance monitoring. Proponents of these types of programs emphasize the potential to identify unanticipated problems. For example, the omnibus surveillance of multiple species throughout a region may identify significant population changes for a particular species that are unexpected and perhaps counterintuitive. These changes would then be a starting point for more intensive monitoring and future hypotheses aimed at identifying the magnitude and causes of these changes. Before reaching this stage, however, surveillance monitoring is arguably necessary. It would be difficult to develop adequate hypotheses for a program that monitors the patterns and changes of the hundreds of bird populations throughout the United States, as the USGS Breeding Bird Survey does (Sauer et al. 2006).

In general, targeted monitoring puts less emphasis on finding and estimating population trends and a greater emphasis on monitoring priority species based on taxonomic status, endemism, sensitivity to threats, immediacy of threats, public interest, and other factors (Elzinga et al. 2001, Yoccoz et al. 2001, Nichols and Williams 2006). Targeted monitoring avoids the largest potential pitfall of surveillance monitoring: that significant parameters are missed because they were not identified early in the planning process. In most scenarios that warrant the implementation of a monitoring program, more specific parameters are integral to attaining the goals and objectives. Therefore, we typically advocate the use of targeted monitoring and the development of clear models, hypotheses and objectives it entails.

Incorporating Stakeholder Objectives

Once the concept of monitoring is clearly defined, you can begin to explore the matter of which monitoring priorities and objectives are best for your program.   Both are often influenced by multiple stakeholders who bring to the table their own goals, needs, assumptions, and predictions that can conflict, coincide, or be mostly unrelated to yours. One primary goal of any monitoring program, therefore, is to incorporate the concerns and predictions of each party. This can be more art than science and is sometimes a challenge to carry out effectively, but it is certainly an important exercise to perform early in the stages of monitoring. Excluding interested parties from the process may require a re-crafting of monitoring objectives after data have been collected, which could undermine an entire program.

The effort to include multiple stakeholders in the early stages of monitoring could be thought of as a pre-emptive attempt at conflict resolution. Any monitoring program must make several decisions with respect to the objectives, the scale of data collection, and what type of data are to be collected. Your ideas of what final decisions might result from the program are entirely subjective. To set things in stone without considering other points of view can result in serious conflicts with many legal and social implications. Conflicts among stakeholders are common in natural resource management. These conflicts are often the result of differing perceptions, varying interpretations of the law, and self-interests that hold the potential to be reconciled with one another and with scientifically rigorous monitoring (Anderson et al. 2001). Thus, eliciting input from stakeholders in an a priori fashion and attempting to resolve conflicts before they become problems is always recommended. Anderson et al. (1999) suggested a general protocol for conflict resolution in natural resource management that can be easily adapted for incorporating stakeholders in wildlife monitoring:

Participants

A good first step is designating a group of people or committee for identifying stakeholders. Having an unbiased group of people, possibly representing different stakeholder groups, to oversee who to invite to the table leads to greater credibility and transparency. Indeed, the careful consideration of participants may be one of the most important steps in any monitoring plan.

Data

Once a group of stakeholders comes to the table there is likely to be a wide-ranging discussion on what types of data are relevant for monitoring. This is an important step in identifying information needs, assessing the potential costs and feasibility of collecting different data types, and agreeing on important state variables. In addition, stakeholders may already have data in their possession that they would be willing to submit for analysis. If stakeholders are already bringing data to the table, it is advisable that all parties sign a “certification” stating that the data have been checked for errors and come complete with metadata.

Analysis

A lofty goal for any initial discussion on monitoring might include an agreement on what data analyses will or will not be used. Although directions may shift as the analytical process proceeds, an early discussion on potential approaches and important assumptions (e.g., independence, parametric assumptions, and representative sampling) can be extremely useful.

Results

It is important for the stakeholders to agree on the interpretation and reporting of results. In many cases two groups of stakeholders could read the same scientific result and reach two different conclusions with different management implications. A clear understanding of the possible results and their interpretation will avoid confusion in interpretation down the road. Falsely assuming that all stakeholders understand analytical results may lead to the creation of a power hierarchy where those more comfortable with quantitative analysis have greater sway or are dismissed because they do not appreciate the more practical aspects of the monitoring plan.

No Surprise Management

Communication is a key component to any successful collaboration. Changes in project goals, objectives, data acquisition, data analysis, and sampling strategies should be updated to a group of stakeholders on a regular basis. Meetings should occur frequently enough that people can discuss ongoing or unexpected trends, and deliberate. These meetings must be balanced with holding meetings too frequently and simply not discussing anything new. Web site updates and webinars can be a useful way of engaging a large number of stakeholders regularly with less impact on their time.

Identifying Information Needs

Information collected should be designed to answer specific questions at spatial and temporal scales associated with the life history of the species and the scope of the management activities that could affect the species (Vesely et al. 2006). Identifying which factors to measure is usually best understood within a conceptual framework that articulates the inter-relationships among state variables (e.g., number of seedlings), processes influencing those variables (e.g., drought), and the scale of the system of interest (e.g., grassland ecosystem). Initially, such a conceptual model represents a shifting competition of hypotheses regarding the current state of our knowledge of a particular system and target species or communities (Figure 4.2). The development of a conceptual model is intellectually challenging and may take months, but this initial step is critical for developers of the monitoring protocol.

When developing a conceptual model, consider the following:

• It should represent your current understanding of the system that you intend to monitor.
• It should help you understand how the system works. What are the entities that define the structure of the system? What are the key processes? This often yields a narrative model—a concise statement of how you think the system works (i.e., a hypothesis).
• It should describe the state variables. What mechanisms and constraints will be included? Which will be excluded? What assumptions will be made about the system? At what spatial and temporal scales does the system operate? This often results in the construction of a schematic model, perhaps a Forrester diagram (a “box and arrow” model).

The conceptual model should allow the key states or processes that are most likely to be affected by management actions to be identified for monitoring. This will provide a framework for generating hypotheses about how the system works and inform the next step in designing the monitoring program: to develop a set of monitoring objectives that is based on these hypotheses and the results of your stakeholder outreach efforts.

The Anatomy of an Effective Monitoring Objective

Developing a conceptual model and understanding stakeholder values leads to the identification of important state variables and processes from which you can derive a set of effective and well-designed management objectives. The objectives serve as the foundation of the monitoring program. A hastily constructed set of management objectives will ultimately limit the scope and ability of a monitoring program to achieve its goals. A well-constructed set will provide the details for how, when, and who will measure the variables that are necessary for successful monitoring. As part of the larger framework, objectives force critical thinking, identify desired conditions, determine management and alternative management scenarios, provide direction for what and how to monitor, and provide a measure of management success or failure (Elzinga et al. 2001). There are three types of objectives that are pertinent to monitoring (Elzinga et al. 2001, Yoccoz et al. 2001, Pollock et al. 2002):

Scientific Objectives

Scientific objectives are developed to gain a better understanding of system behavior and dynamics (Yoccoz et al. 2001). In this case, a set of a priori hypotheses is developed to predict changes in state variables in response to environmental change. For example, a set of hypotheses regarding the population dynamics of shrubland songbirds in Connecticut may identify several state variables (e.g., bird abundance, presence/absence, reproductive success) and how those variables may change due to changing environmental conditions (e.g., drought, disturbance, land use change). In this case, several hypotheses are generated that readily translate to monitoring objectives. The key to using scientific objectives is to develop competing hypotheses and predictions that can be compared to patterns resulting from data analyses.

Management Objectives

Management objectives incorporate the predicted effects of management actions on system responses. These objectives describe a desired condition, identify appropriate management steps if a condition is or is not met, and provide a measure of success (Elzinga et al. 2001). Not unlike scientific objectives, management objectives should be developed using a priori hypotheses of how a species or population will respond to a given management action. The data collected are then compared to these predictions.

Sampling Objectives

Sampling objectives describe the statistical power that one is attempting to achieve through their management objectives. Many management objectives will seek to estimate the condition and/or a change in a target population (e.g., a 10% increase in juvenile survivorship), but the degree to which that estimate approximates the true condition will, in part, be a function of its statistical power. Consideration of statistical power is critical within a monitoring framework because of the implications of missing a significant effect (Type II error) and not initiating management when it is necessary to do so. In a monitoring program, the perceived condition of a system relates to a target or a threshold in a current state to a desired state. These targets or thresholds are reflected in management objectives. For example, a threshold objective would be limiting the coverage of a wetland site by an invasive species such as common reed to less than 20%. Once that condition (i.e.,< 20%) is reached or exceeded, a management action would be initiated. Management objectives may also relate to an active change in an existing state. That is, a change objective would be decreasing the cover of Common reed to less than 10%. In both of these cases, a sampling objective describes the statistical precision and variation associated with estimating that condition or change, oftentimes using a confidence level (e.g., be 90% confident the coverage of Common reed is estimated to within ± 5%). Sampling objectives relate to the statistical power of your sampling scheme and better inform you about your power to detect significant change (Gibbs et al. 1998).

Whether it is a scientific, management, or sampling objective, all monitoring objectives should consist of several key ingredients (Figure 4.3):

What?

The most important component of a monitoring objective identifies what will be monitored. Will you be monitoring a species or a group of species? Is the focus a specific population? Is the focal species an indicator species serving as a surrogate for another species or habitat type? Also, be clear as to the parameter associated with this target that you will be measuring. Are you collecting information on abundance, occurrence, reproductive success, demographics, or density?

Where?

The site or geographic area should be clearly delineated. Often times, managers cannot apply a monitoring program over the entire area of interest. In these cases, sample areas must be specified and the results of these sampling areas used to draw inference to the rest of the study area (Pollock et al. 2002). It is important that these sampling areas be defined objectively, be representative of the larger study area, be free of biases, and represent an appropriate spatial scale relative to the species or processes of interest.

When?

Providing a time frame is critical for achieving a monitoring objective. Time frames should incorporate species’ life history characteristics (e.g., breeding season, flowering, longevity), logistical constraints, and political schedules. Short time frames are always preferred due to changes in budget, management adaptation, and the uncovering of unexpected information (Elzinga et al. 2001), but may not be useful if dealing with long-lived organisms. When monitoring programs must be necessarily conducted over long time periods, considerable attention should be given to the likelihood of continued funding to support the program.

Who?

Often overlooked, the earlier one can identify who is conducting the monitoring the more likely that the program will be implemented correctly. This avoids the inevitable ambiguity of “passing of the buck”. In addition to the responsibility of conducting the monitoring efforts, is the involvement of stakeholders in the design and interpretation of the data. When continued funding is needed to maintain a monitoring program over long periods of time, key stakeholders include those involved in budgeting. At all stages of monitoring program managers should continually be thinking of who in our society could be affected by these results and ensure that they are kept informed.

Articulating the Scales of Population Monitoring

Another important aspect of setting goals and objectives is to define the scales of space and time over which monitoring should occur. Most populations occupy a landscape representing different habitat patches of varying quality. As a result, individuals in a population are interdependent on a number of areas that are influenced by constantly changing environmental conditions. Population trends are not only dependent on the quality of individual patches and areas, but also by the spatial and temporal distribution of suitable and unsuitable habitat patches. Management and development can have an immediate impact on the accessibility of habitat for a species, but the impact can also be delayed if the habitat changed is utilized seasonally or only in a particular ecological context, or if the manager or developer incorporates a species’ needs into her endeavors. Consequently, a hierarchical approach that takes a broad perspective that is not restricted by politically- and management-imposed boundaries, and allows an examination of population and habitat change across multiple scales should be considered when developing a monitoring plan.

Project or Site Scale

Managers and regulators often want to know what effect a specific management action will have on a population. Indeed, many of the factors that influence populations occur at the site level. For forests this might be a stand, for a town it could be an individual property, or for agriculture, a field. Factors such as resource availability, predation, parasitism, and competition that change as a result of management at these local scales can affect a number of demographic processes including fertility, survivorship, mortality, and dispersal within focal populations. Although these processes are important, the ability to repeatedly monitor them at the site level can be logistically and economically difficult. Even when it is possible, however, the most critical question that pertains to studying of species in small areas must still be addressed: Are the changes that are documented at the site-level are reflective of those occurring in the same species over a wider geographic area? For species with a restricted geographic range such as some plants and fish stocks, this scale may be an accurate representation of population change. In general, if the species is influenced by the same factors (e.g., weather, resources, predation, topography) throughout its geographic range, then fluctuations on a small scale will often accurately represent fluctuations on a wider scale. However, if factors are site-specific, as they often are, then fluctuations in occurrence and abundance will vary from place to place (Holmes and Sherry 1988). In this case, monitoring population change at a larger scale will be more meaningful. It is important to note, however, that as the scale gets larger, identifying the causes of any observed changes can be increasingly difficult.

Landscape Scale

Since many animal populations are dynamic, occupy a heterogeneous landscape, and use multiple resources across that landscape, it is important that population monitoring incorporates the context of any particular site. Indeed, monitoring efforts must incorporate areas surrounding the forest stand, field or other site, to include water bodies, ownership, other habitat types, barriers and corridors. This is especially important for species that have large home ranges and/or use multiple sites during their life history, or for species that depend on dynamic habitats that are influenced by weather, season, or succession.

The arrangement of resources over a large geographic area can be critical for sustaining a population. Monitoring over large complex landscapes may increase your ability to detect long-term changes in abundances, predation rates, extinction rates, patch dynamics, metapopulation dynamics, disturbance effects, and rates of human influences on focal species (Noss 1990). However, monitoring at this scale also presents difficulties in design and logistics, because landscapes, not patches, become the sampling units (McGarigal and McComb 1995, McGarigal and Cushman 2002, Meffe et al. 2002, Pollock et al. 2002).

At the landscape scale, population trends are not only dependent on the quality or extent of individual habitat patches, but also on the spatial distribution, pattern, and connectivity of suitable and unsuitable patches (Meffe and Carroll 1997). Landscape pattern, in turn, is influenced by the type of patches, size of patches, length of surrounding edge, barriers between patches, and nature of corridors (McComb 2001, McComb 2007). Fragmentation, for instance, can lead to the isolation of some important habitat patch types, which can lead to disruptions in species dispersal, and eventually to extinction despite management to conserve a species at a site scale. Whether an area is connected or fragmented ultimately depends on the habitat requirements and dispersal abilities of the organism. Nonetheless, at times even sampling over large landscapes cannot answer key questions because no information is provided concerning the effect of the context of the landscape on a species’ population dynamics. Consequently, sampling at even larger spatial scales may be needed for some species.                         

Range-Wide Scale

A larger perspective is especially important for sensitive and management-indicator species because many have developed morphological and behavioral adaptations that are unique to certain geographic locations (Cody 1985). Range-wide data and trends provide a perspective into localized population changes. For example, populations of Blue-winged Warblers have been declining in Connecticut at a rate of 3.4% per year over the last 40 years (Figure 4.4). However, the concern over this rate of decline is greatly amplified because an estimated 13% of the global population of this species is found in Connecticut (Figure 4.5, Rosenberg and Wells 1995).

There are few examples of successful long-term monitoring programs that document changes in populations throughout entire geographic ranges, but the North American Breeding Bird Survey, Waterfowl Harvest Surveys, Salmon Escapement monitoring, USDA Forest Service’s Forest Health Monitoring Program, and Agricultural Production monitoring have produced useful results. Ideally, all monitoring protocols would be designed to detect population changes in a species throughout its range; however, this is often logistically and financially prohibitive. This is particularly true for the many species that may have broad geographic ranges. Yet even if active monitoring cannot be undertaken on a range-wide scale, a monitoring program must make an attempt to understand the context of any population changes it documents in the most informed way possible.

Organism-centered Perspective

The most important spatial scale to a species is defined by its life history and habitat requirements. The different habitat requirements between groups of birds, insects, and mammals are clear, but even within these larger groups, species that share similar preferences for food, water, and cover, may have vastly different requirements for space. Species with similar requirements for vegetation and other resources are often affected differently by management actions due to differences in area-sensitivity and home range sizes. An organism-centered view suggests that there is no general definition or perspective of habitat pattern. Therefore the effects of management actions that alter the spatial arrangement of habitat patches across a landscape will ultimately vary from one species to the next. For example, a 10-ha clearcut can have a profound influence on a small forest passerine with a restricted territory size, such as a Black-and-White Warbler, but have relatively no impact on a Cooper’s Hawk whose territory includes hundreds of hectares.

Area requirements and sensitivity to patch size are not the only factors influencing a species’ response to management. The dispersal capabilities of species are a critical component that determines whether or not a population is directly effected by loss of habitat resulting from management or, inordinately affected by fragmentation of its habitat . Species migrate between habitats that are separated by ecological and anthropogenic barriers. However, each species differs in its perception of these “gaps” in habitat and therefore in its ability to successfully cross them (With 1999). A landscape is fragmented if individuals cannot move from patch to patch and are isolated within a single area. Simulations have suggested that species with limited dispersal capabilities are much less likely to successfully cross habitat “gaps” to other habitat clusters relative to species with a higher ability to disperse (With 1999) (Figure 4.6). Whether or not management causes fragmentation, and how monitoring should address this effect, must be addressed from an organismal perspective.

Data Collected to Meet the Objectives

After creating your conceptual model of population persistence for your species and putting this into the proper spatial context, specific questions regarding the potential impacts of management on a species should emerge. The scope of the monitoring program should lead the investigators to identify a set of questions that can be addressed by different data types. For instance, consider the following five questions and the decisions that could be made to meet the information needs associated with each (from Vesely et al. 2006):

1. Given our lack of knowledge of the distribution of a clonal plant species, we are concerned that timber management plans could have a direct impact on remaining populations that have not yet been identified on our management district. How will we know if a timber sale will impact this species?

In this example the plant species may have a geographic range extending well beyond the timber sale boundaries and over multiple National Forests, but populations of this species are patchily distributed and their abundance is poorly known. Based on our conceptual model of persistence, therefore, we are concerned that population expansion and persistence may be highly dependent on movement of propagules among sub-populations and that additional loss of existing patches may exacerbate the loss of the population over a significant portion of its range. Consequently the primary goal of a monitoring effort should be to identify the probability of occurrence of the species in a timber sale. A survey of all (or a random sample of) impending timber sales will provide the land manager with additional information with regard to the distribution of the species. Although information may be collected that is related to fitness of the clone (size, number of propagules, etc.), the primary information need is to estimate the probability of occurrence of the organism prior to and following management actions. Indeed, this survey and manage approach also lends itself well to development of a secondary monitoring approach that utilizes a manipulative experiment. Explained in more detail later, identification of sites where the species occurs can provide the opportunity for random assignment of manipulations and control areas to understand the effect of management on the persistence of the species. This approach may be particularly important when dealing with species, such as cryptic or infrequently apparent species, where the probability of not detecting individuals is relatively high despite the species being present on the site.

2. Given the uncertainty in the distribution of a species of a small mammal species over a management area, we are concerned that planned a timber harvest could have an undue impact on a large proportion of individuals of this species within the management area. We need to have an unbiased estimate of the abundance of the species over the entire planning area to understand if the proposed management activities indeed have the potential to impact a significant portion of the population.

In this example, the species geographic range extends well beyond the boundaries of the management area, but the manager is concerned that the sites under consideration for management may be particularly important for the species’ persistence within her management area. The manager therefore needs to understand the dynamics of the population within the sites, but also the population dynamics within the entire management area, as well as the interplay between these levels to understand the full potential for adverse effects on the species. Based on survey information it is clear that the species occurs in areas that are planned for harvest. But do they occur elsewhere in the management area? With an unbiased estimate of abundance, that extends over the area (or forest, or watershed, etc.) one can estimate (with known levels of confidence) if the proposed management activities might affect 1% of the habitat or population for this species or 80% of the habitat or population. Consider the differences in management direction given these two outcomes. Collecting inventory information following standardized protocols over management units provides the manager with a context for proposed management actions and is integral to a successful hierarchical approach.

3. Given the history of land management on a Refuge, how will the future management actions described by the current Comprehensive Conservation Plan (CCP) influence the abundance and distribution of a sub-population of a salamander species that we know occurs on our Refuge?

In this example, the species again has a geographic range that extends well beyond the boundaries of the Refuge, but there is concern that the relatively immobile nature of subpopulations of this species may make the animals on the Refuge highly important in contributing to its range-wide persistence. If the Refuge’s subpopulation is adversely impacted by management and has shown historical declines in abundance as a result of the past management activities, the CCP may need to be amended. The goal, therefore, is to establish the current status of the species on the Refuge and allow managers to detect trends in abundance over time. Changes in abundance or even occurrence may be difficult to detect at the project scale (e.g, road building), because individuals are patchily distributed, but if data are collected cumulatively over space and time, impacts could become apparent. Consequently, this status and trends monitoring approach should extend over that portion of the Refuge where the species is known or likely to occur and provide an estimate of abundance of the species at that scale at several different time periods. It is important to distinguish between an estimate of abundance over a large area (inventory), and a total count of all individuals in an area (census). Inventories, when conducted following sampling guidelines, and accounting for detection probabilities, can produce estimates with known levels of confidence. Censuses often are not cost effective unless the species occurs in very low numbers and the risk of regional or range-wide extinction is high.

In short, the focus of this monitoring effort would be to document changes in abundance over time over a spatial scale that encompasses the sub-population of concern. One final consideration is that, if at all possible, the abundance estimates should be specific to age and sex cohorts to allow managers to identify potential impacts on population demographics. For instance, reduction in the oldest or youngest age classes, or of females, may provide information on recruitment rates that is significant enough to cause changes in management actions before a significant change in total abundance occurs.

4. Assume that concern has been expressed for a species of neotropical migrant bird whose geographic range extends across an ecoregion. The monitoring plan needs to assess if the history of land management throughout the ecoregion and the multiple plans for future management applicable to the region are contributing to changes in populations over time. In other words, are multiple types of management having an effect on the population?

In this example we are dealing with a species that is probably widely distributed, reasonably long-lived, and spends only a portion of its life in the area affected by proposed management. One could develop a status and trends monitoring framework for this species, but the data resulting from that effort would only indicate an association (or not) with time. It would not allow the manager to understand the cause and effect relationship between populations and management actions.

In this case there are several strata that must be identified relative to the management actions. Can the ecoregion be stratified into portions that will not receive management and others that will receive management? If so, then are the areas in each stratum sufficiently large to monitor abundance of those portions of the populations over time? Monitoring populations in both strata prior to and following management actions imposed within one of the strata would allow the managers to understand if changes occur in the most important response variables from the conceptual model due to management. For instance, if populations in both managed and unmanaged areas declined over time, then the managers might conclude that population change is independent of any management effects and some larger pervasive factor is leading to decline (e.g., climate change, changes in habitat on wintering grounds). On the other hand, should populations in the unmanaged stratum change at a rate different from that on the managed stratum, then the difference could be caused by management actions, and lead managers to change their plan.

5. Finally, say that the conceptual model suggests that the most likely factor affecting the change in population of a wide-ranging raptor is nest site availability. At the ecoregion scale, population density is low and the probability of detecting a change in abundance or fitness at that scale is likewise very low. Rather, managers may wish to monitor habitat elements that are associated with demographic characteristics of the species. How might a monitoring protocol be developed that would allow managers to use habitat elements as an indicator of the capability of an ecoregion to contribute to population persistence?

An unbiased estimate of the availability of habitat elements assumed to be associated with a demographic characteristic of the species and an estimate of the demographic characteristic assumed to be associated with the habitat elements are needed to develop wildlife habitat relationships. Ideally, monitoring of the habitat elements and the demographic processes can be conducted to assess cause and effect relationships (see above), but with rare or wide-ranging species this may not be possible. In these cases, testing a range of relationships through use of information theoretic approaches can help you to identify the ‘best’ relationship given the limitations of the data (Burnham and Anderson 2002). Regardless of the resulting monitoring design, it is important that the monitoring framework for the vegetation component of the habitat relationship is implemented at spatial and temporal scales consistent with those used by the species of interest.

Which Species Should Be Monitored?

If you have several options for species or groups of species that, if monitored, will yield data that meets your objectives, how should you decide which to monitor? Where should you focus time and money? Although the species selected will oftentimes be driven by the values of the stakeholders associated with land use and land management in the area of interest, sometimes characteristics of the species themselves help to focus the list. The following categories of species are some of those most commonly viewed as worthy of special consideration and therefore particularly useful for practitioners when selecting the species to be monitored:

1. Level of Risk – the perceived or real level of risk of loss of the species from the area now and into the future. Risk can be based on previously collected data, expert opnion and stakeholder perceptions.
2. Regulatory status – Species listed under State and/or Federal threatened or endangered species legislation.
3. Government Rare Species or Communities classification –those species or plant communities designated by federal or state agencies as in need of special consideration.
4. Restricted to specific seral stages – species sensitive to loss of a vegetative condition such as a stage of forest, wetland, or grassland succession. Species associated with seral stages or plant communities that are under-represented relative to a reference condition or the historic range of variability often rise to the top when identifying focal species.
5. Sensitivity to environmental change/gradients – species sensitive to environmental gradients such as distance from water, altitude, soil conditions, or characteristics. Under current climate change scenarios, for instance, species associated with high altitudes or high latitudes are of particular concern.
6. Ecological function – species that are particularly important in modifying the processes and functions of an ecosystem. For instance, gophers expose soil in grasslands and voles move mycorrhizal fungal spores in forests.
7. Keystone species — species whose effects on one or more critical ecological processes or on biological diversity are much greater than would be predicted from their abundance or biomass (e.g., beaver, large herbivores, predators).
8. Umbrella species — species whose habitat requirements encompass those of many other species. Examples include species with large area requirements or those that need multiple vegetative conditions, such as raptors, bears, elephants or caribou.
9. Link species — species that play critical roles in the transfer of matter and energy across trophic levels or provide a critical link for energy transfer in complex food webs (e.g., insectivorous birds) or which through their actions influence trophic cascades effects (Ripple and Beschta 2008)
10. Game species – species that are valued by segments of society for recreational harvest.
11. Those for which we have limited data or knowledge – monitoring may provide an information base necessary to understand if continued monitoring is needed.
12. Public/regulatory interest – some species are simply of high interest to the general public because of public involvement (e.g., bluebirds, wood ducks, rattlesnakes). These can include species that are desirable as well as those that interfere with people’s lives.

Intended Users of Monitoring Plans

Monitoring plans can be useful for a variety of users including agency managers and planners, the general public, politicians (to ensure adherence to local, state and federal legislation), non-governmental organizations with similar missions, and also industries with Habitat Conservation Plans on adjacent or nearby lands. Different components of a monitoring plan often are useful to different stakeholders. For instance, a survey prior to a management action may allow a manager to alter the management action to accommodate a species found on the site during the survey. A declining trend in a focal species population in the managed portion of a site compared to an unmanaged portion may allow a Land Planner to make changes in an adaptive management framework over the site. While at a larger scale, regional declines in a species on public land-holdings may lead to legislation or agreements that span ownership boundaries across the species’ geographic range to encourage recovery. Expected products will be dependent on the questions that are asked. Occurrence, abundance, fitness, range expansion/contraction, each may be appropriate to address certain questions. Whatever the measure, whatever the question, and whatever the expected product, the results must be effectively communicated so that the manager, planner or politician can make an informed decision regarding the likely effects of a management action or legislation on the long-term persistence of the species.

Summary

Some biologists distinguish between targeted and surveillance monitoring. Targeted monitoring requires that the monitoring design and implementation be based on a priori hypotheses and conceptual models of the system of interest. Surveillance monitoring oftentimes lacks hypotheses, models or specific objectives. The structure of each approach is driven largely by societal values. Stakeholder involvement in identification of indicators and thresholds for changes in management efforts is a key step in developing a monitoring plan. Suggested steps in stakeholder involvement include:

• Identify the participants.
• Agree on the types of data needed.
• Agree on the types of analysis to be used.
• Agree on how the results will be interpreted.
• Agree on ‘no surprises’ management

Developing a conceptual model and understanding stakeholder values can help to identify important state variables and processes from which management objectives will emerge. Objectives should consider the following questions: What, Where, When, and Who? Objectives also have a scale component. Deciding if the results will address questions at the project, landscape, or geographic range scales influences the utility of the information that is gathered. Alternatively monitoring organisms may be most appropriate. Finally where there is a choice as to which species to monitor, the values of the stakeholders may guide species selections. Rare species invariably rise to the top of a list, but economically important species, keystone species or species that are indicative of ecosystem stresses may also be selected depending on the stakeholder interests.

References

Anderson, D.R., K.P. Burnham, A.B. Franklin, R.J. Gutiérrez, E.D. Forsman, R.G. Anthony, and T.M. Shenk. 1999. A protocol for conflict resolution in analyzing empirical data related to natural resource controversies. Wildlife Society Bulletin 27:1050–1058.

Anderson, D. R., K. P. Burnham, and G. C. White. 2001. Kullback-Leibler information in resolving natural resource conflicts when definitive data exist. Wildlife Society Bulletin 29:1260-1270.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and inference: a practical information-theoretic approach. Springer-Verlag, New York, New York, USA.

Cody, M.L. 1985. Habitat selection in birds. Academic Press, Orlando, Florida.

Elzinga, C.L., D.W. Salzer, and J.W. Willoughby. 1998. Measuring and monitoring plant plant populations. Technical Reference 1730-1. Bureau of Land Management, National Business Center.

Elzinga, C.L., D.W. Salzer, J.W. Willoughby, and J.P. Gibbs. 2001. Monitoring plant and animal populations. Blackwell Science, Inc., Malden, Massachusetts.

Gibbs, J.P., S. Droege, and P. Eagle. 1998. Monitoring populations of plants and animals. Bioscience 48:935-940.

Holmes, R.T., and T.W. Sherry. 1988 Assessing population trends of New Hampshire forest birds: local and regional patterns. Auk 105:756-768

McComb, B.C. 2007. Wildlife habitat management : concepts and applications in forestry. CRC Press/Taylor & Francis Group, Boca Raton.

McComb, W.C. 2001. Management of within-stand features in forested habitats. Pages 140-153 in D.H. Johnson, and T.A. O’Neil, editors. Wildlife habitat relationships in Oregon and Washington. Oregon State University Press,Corvallis, Oregon.

McGarigal, K., and S.A. Cushman. 2002. Comparative evaluation of experimental approaches to the study of habitat fragmentation effects. Ecological Applications 12:335-345.

McGarigal, K., and W.C. McComb. 1995. Relationships between landscape structure and breeding birds in the Oregon Coast Range. Ecological Monographs 65:235-260.

Meffe, G.K., L.A. Nielsen, R.L. Knight, and D.A. Schenborn. 2002. Ecosystem management: adaptive, community-based conservation. Island Press, Washington, D.C.

Meffe, G.K., and C.R. Carroll. 1997. Principles of conservation biology. 2nd edition. Sinauer, Sunderland, MA.

Nichols, J.D. 2000. Monitoring is not enough: on the need for a model-based approach to migratory bird management. Pages 121-123 in Proceedings of Strategies for Bird Conservation: The Partners in Flight Planning Process. Proceedings RMRS-P-16:121-123.

Nichols, J.D., and B.K. Williams. 2006. Monitoring for conservation. Trends in Ecology and Evolution 21:668-673.

Noss, R.F. 1990. Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology 4:55-364.

Pollock, K.H., J.D. Nichols, T.R. Simons, G.L. Farnsworth, L.L. Bailey, and J.R. Sauer. 2002. Large scale wildlife monitoring studies: statistical methods for design and analysis. Environmetrics 13:105-119.

Ripple, W.J. and Beschta, R.L. 2008. Trophic cascades involving cougar, mule deer, and black oaks in Yosemite National Park. Biological Conservation. 141: 1249-1256.

Rosenberg, K.V., and J.V. Wells. 1995. Important geographic areas to Neotropical migrant birds in the Northeast. Report to United States Fish and Wildlife Service, Hadley, Massachusetts.

Sauer, J.R., J.E. Hines, and J. Fallon. 2006. The North American Breeding Bird Survey, Results and Analysis 1966 – 2006. Version 6.2.2006. USGS Patuxent Wildlife Research Center, Laurel, MD

Vesely, D., B.C. McComb, C.D. Vojta, L.H. Suring, J.Halaj, R.S. Holthausen, B. Zuckerberg, and P.M. Manley. 2006. Development of Protocols to Inventory or Monitor Wildlife, Fish, or Rare Plants. U.S. Department of Agriculture, Forest Service.

Whitman, A.A., and J.M. Hagan. 2003. Biodiversity Indicators for Sustainable Forestry. National Center for Science and the Environment. Washington DC.

With, K.A. 1999. Is landscape connectivty necessary and sufficient for wildlife management? Pages 97-115 in J.A. Rochelle, A. Lehmann, and J. Wisniewski, editors. Forest Fragmentation: wildlife and management implications. Koninklijke Brill, Leiden, The Netherlands.

Yoccoz, N.G., J.D. Nichols, and T. Boulinier. 2001. Monitoring of biological diversity in space and time. Trends in Ecology & Evolution 16:446-453.