Ecosystem Management Tools and Techniques:
Proceedings of a CRS Workshop

94-430 SPR

APPENDIX I. PRESENTATIONS AT THE WORKSHOP

Introduction to the CRS Workshop on the Tools and Techniques of Ecosystem Management
The Role of Data Management and Information Analysis for Supporting an Ecosystem Approach
Applications of GPS and GIS Techniques in Migratory Bird Management

INTRODUCTION TO THE CRS WORKSHOP ON THE
TOOLS AND TECHNIQUES OF
ECOSYSTEM MANAGEMENT

Wayne A. Morrissey, Moderator (4)

Good morning, I am Wayne Morrissey of the Science Policy Research Division of the Congressional Research Service (CRS), and moderator of the CRS workshop on the Tools and Techniques of Ecosystem Management. I would like to take this opportunity to welcome our audience and special guests who have been, and will be, demonstrating the tools and techniques of ecosystem management this morning.

Some experts have described ecosystem management as, and I quote, "An emerging and evolving paradigm which consists of a number of similar processes that seek ecological approaches for the sustainable management of the environment, including social and cultural values, and economic interests."

This workshop is the first in a series which follows the CRS Symposium on Ecosystem Management held last March. At that symposium, 150 experts including scientists, natural resources managers, Federal, State and local government representatives, and private land owners, met to discuss opportunities and concerns about developing a Federal ecosystem management policy.

A series of up to 6 CRS workshops to be held over the next year will address ecosystem management topics. Today our intention is to demonstrate that ecosystem management does not merely exist as a concept, a hope, or an aspiration, but that it has been, can be, and is actually being, done right now. However, some might concede that existing approaches are not as comprehensive or sophisticated as they will need to be to enable successful ecosystem-based management approaches.

Today, we hope to develop better understanding from those actually involved how ecosystem management, from a technological perspective, is or isn't working. We hope to learn about the potential opportunities and limitations of these tools; and, moreover, those questions relating to ecosystem management that they can and cannot answer at this time.

We have a distinguished team of experts with us today. I say team, because it will soon become clear that another purpose of today's workshop is to demonstrate how State and local governments, industry and academia are working together with the federal government both to develop the tools and techniques of ecosystem management, and to actively participate in training and research in the field.

Another theme that will also become apparent, before we finish today's workshop, is that government-wide activities in ecosystem management, and especially the data gathering and data management components of those activities, will have implications for broader national goals such as conducting the National Biological Survey, implementing Federal natural resources planning and management programs, and developing a National Spatial Data Infrastructure (NSDI), among others.

It is with great pleasure that I introduce today's guests. In the order that they will give their presentations; they are:

• Nancy Tosta, Chief of the Branch of Geographical Data Coordination, in the National Mapping Division of the U.S. Geological Survey, in Reston, Virginia;

• D. Alan Davenport of the U.S. Fish and Wildlife Service, Biologist and Geographic Information System (GIS) specialist in Laurel, Maryland;

• Dr. Mike Scott who is working for the National Biological Service, and is Leader for the Idaho Cooperative Fish and Wildlife Research Unit at the University of Idaho, in Moscow, Idaho. Dr. Scott is accompanied by Brian Biggs of the U.S. Geological Survey's Eros Data Center in Sioux Falls, South Dakota, who is also working with the United Nation Environmental Programme's Geographical Resources Information Data (GRID) program; and

• Virginia Ferreira of the Terrestrial Ecosystems Regional Research and Analysis Laboratory (TERRA LAB) of Fort Collins, Colorado. Ms. Ferreira is accompanied by a cadre of assistants whom she will introduce later.

In the audience, we also have members of the Federal Interagency Ecosystem Management Coordinating Group (IEMCG), who will be available for questions and discussion when concurrent demonstrations resume after the presentations.

CRS would also like to thank Barry Gold of the House Subcommittee on Technology, Environment, and Aviation and Gloria Dunderman of the House Committee on Science, Space, and Technology for their assistance in co-hosting this event on behalf of the Honorable Tim Valentine, Chairman of the Subcommittee, who requested CRS sponsor this workshop.

I would also like to thank Dr. Jeffrey Zinn of the CRS Environment and Natural Resources Policy Division for co-coordinating this workshop. Dr. Zinn was also a coordinator of the overall CRS Symposium on Ecosystem Management. Tom Miller, of the CRS Programs Office, was especially helpful with the logistics of pulling this workshop together. Several others at CRS who also work on various aspects of ecosystem management are here today. They include, Dr. M. Lynne Corn of the Environment and Natural Resources Policy Division, and John Justus and Michael Simpson of the Science Policy Research Division.

And, most of all, I would like to thank you, the audience, for coming to learn about the tools and techniques of ecosystem management. We want you to ask serious questions of our guests after their individual presentations, but we also want you to have fun. you shouldn't feel guilty that you are having fun at this workshop. If you aren't having fun already, I'm sure you will, when you have a chance to view the demonstrations that our speakers will describe, and when TERRA Lab invites you to participate in their hands-on "Active Response GIS," later in the program.

Although few, if any, would concede that ecosystem management is not a matter of serious business, I'm sure that many of the professionals who are here demonstrating and discussing these technologies today also enjoy their work and their fields.

THE ROLE OF DATA MANAGEMENT AND INFORMATION
ANALYSIS FOR SUPPORTING AN ECOSYSTEM APPROACH

Presented by Nancy Tosta (5)

There are a number of people who are likely to give you more technical perspectives on tools and techniques for ecosystem management. I am going to focus on the philosophical aspects--what we are trying to do related to coordination among institutions interested in spatial data, and how we think about the places we are trying to manage. I used to be a technologist, but now I am primarily a coordinator.

Rather than discussing the tools for ecosystem management, which I think about as being shovels, stream gauges, or controlled beams, for example, we are really addressing tools and techniques for ecosystem "information management." I would like to clarify what we in the geospatial data community are talking about when we refer to "ecosystems" and "information". E.O. Wilson's definition of biodiversity references ecosystems. He states that they (ecosystems) "comprise both the communities of organisms within particular habitats and the physical conditions under which they live." He makes it clear that it is not just the organisms, but it is very much the physical environment that impinges upon them. The physical environment relates directly to much of the work that I do in the geospatial data community, but we will come back to that.

Our concept of information is changing in the digital world. A March 1994 article in Wired magazine, which deals with information technology and life on the "net" (Internet), by John Perry Barlow entitled the "Economy of Ideas," examines the issues that are going to be associated with trying to copyright digital data. Barlow also provides some perspectives on information which I have found to be thought-provoking as I deal with data and technology information. He states that there are the three characteristics of information: it is an activity, a life form, and a relationship. Basically, if you think of information as an activity, you must imagine it as a verb, not a noun. He describes it as the pitch, not the ball. It is the interaction of things, of ideas with things. Data have to interact with your mind for them to be information. It is experienced, and you do not own it. you still have it, even if you give it away. It is necessarily distributed; it definitely has a half-life . It has to move. The notion of hoarding information is almost an oxymoron. In a sense information is a life form in that it definitely wants to be free; it wants to be passed around; it wants to be used; it replicates into the cracks of possibility.

Anyone who has seen a rumor spread and evolve knows that information wants to change. I have been amazed recently tracking a couple of statements that I was present at the source of and finding whole new thoughts coming back to me two weeks later. Information certainly does have a half-life. There is a lot of relevance in many cases to the topics to be addressed in this workshop.

Barlow describes information as a relationship: If you think about this, you realize that the meaning of any information is unique. What you think is valuable information may not necessarily be valuable to anyone else. you have a personal relationship with information. That is one of the reasons organizations struggle over pricing data. The value of information often depends on use and context. Familiarity has value. If you are comfortable with the source of the information, it may mean more to you. There may be certain places that you are comfortable getting information from and others that you are not.

The same is true with the point of view--time, place, or space--it may be that information is more valuable because it is current rather than necessarily represents some known geography but is dated. And information is its own reward. Information often begets more information when you use it. These thoughts seem worth contemplating as we consider dealing with the information economy and information as a currency. I like to play with these ideas as we try to put policies in place that relate to the management of information.

With ecosystems, as with anything real that needs to be studied or managed, we observe and we measure. We create data when we take observations and measurements. Those data we organize and integrate, and hopefully they become information. We can interpret and understand that information to gain knowledge, and, hopefully, we can actually use that knowledge intelligently to manage the real things we were measuring -- and that in the process we might consider ourselves as gaining wisdom. One of the technology tools that we have been working with to manage ecosystem data is the Geographic Information System (GIS). GIS helps us organize data so that it becomes information. In the transformation of information to knowledge, values play a major role in how we understand it, how we interpret our point of view, and what we hold to be truth; although we don't often talk about this. I was interested to see some of the work that TERRA Lab is doing with collaborative studies, because as we try to translate our understanding of information into a basis for making decisions we must consider all of the perspectives.

Ecosystem measurements and observations fall into categories. There are the living pieces of ecosystems, such as the biological organisms, the habitats, plant associations, and vegetation. There is the physical or geographical environment, including terrain, soils, geology, the transportation infrastructure, and hydrology. There are environmental measurements that we take that often look at specific problems such as pollution or toxic sites. There are political and institutional aspects of ecosystems, such as the administrative boundaries, ownership, the basis of the economy in the ecosystem, and how the economy affects how the ecosystem is managed. And certainly there are cultural aspects of ecosystems. Aside from wilderness areas, there are relatively few ecosystems any more that are not somehow tied up with human culture. All of these variables and factors must be considered to conduct ecosystem studies. The tools have to give us the ability to integrate the representations, measurements, and various values for these factors.

Another challenge with ecosystem information is that it is multi-resolutional in terms of both time and space. The kinds of measurements that must be taken must recognize that things change at different rates and must be measured at different rates. We take a human population census only once every 10 years on a national basis, although we know the population changes every minute. We know that the rates of change of certain things vary, but we often only have the resources to measure them at certain times ; and I am not sure we yet have good tools to track variable rates of change. Ecosystems obviously vary spatially as well, from something that is very small--under a log or in a pond--to concepts such as the redwood belt and bioregions. How do we integrate these variable measurements and scale to yield meaningful results?

I mentioned earlier that GIS has become a very popular tool. you see variations of GIS around the room, and will hear more about them from other speakers. GIS has become a way to organize and integrate observations and measurements using location as the common denominator. Many of the origins of GIS have come from mapping and traditional map products. We are seeing an evolution away from that. Rather than maps being the primary input into GIS or spatial data models, we are seeing measurements, real time activities actually, being what we want to feed our GIS. Maps are still a primary output--the way to display the results of analyses, but we are increasingly using other data sources as the input. We also link many models, including process models, statistical models, and visualization techniques, to GIS. And we are beginning to think less of GIS as an application tool, and more as an information organizing tool. It gives us the ability to relate events, and activities, and things that occur in a common place. It may be that when you look at an ecosystem, perhaps you have multiple organisms that have been inventoried. Associated with these are a variety of taxonomic classifications that are not really linked in space, but are related by the organism. The occurrences of these classifications may be related, and defined as spatial attributes, such that future analyses can begin to show distributions of related species. GIS thus becomes a powerful organizing, as well as analysis tool.

We have come to call such data geospatial, and we are working on something called the national spatial data infrastructure. When I say "we", I refer to the entity called the Federal Geographic Data Committee. Geospatial data, as defined in an Executive Order signed by President Clinton on April 11, 1994, are any information related to a location on the surface of the earth. Obviously, geospatial data have a lot to do with ecosystem management.

The Federal government is spending considerable funds collecting geospatial data. Last year it conducted a data call asking Federal agencies how much they were spending on the collection, management, and dissemination of geospatial data. Using the definition I just mentioned, the total exceeded $4 billion annually. We suspect this is low because it does not include all of the image activities, a lot of the sensor activities, climatological data, and so on. These data are often collected in a stovepipe approach. They are mission driven; collected by individual agencies carrying out functional responsibilities, and often over the same piece of geography. Data bases are built and analyzed, and policy recommendations are made in isolation from other agencies and levels of government that also might share an interest in that same piece of land.

Often, these data are not made readily available, or they are hard to find. The Yellow Pages do not have a listing of all the data sets that might be relevant in a specific geographic area. Even if found, these data can be difficult to access; obtaining a piece of a file with the specific measurements of interest may be costly or time-consuming. There are often questions of copyrights and data fees. This is not as true among Federal agencies, but State and local governments often set such policies. In many cases, other entities will take Federal data, build on it, add "value" to it, and then copyright it.

Over the last couple of years, data revenues have been debated as a source of funding for data collection and maintenance programs. We are going to see a lot of on-going debate in this area as data sharing becomes more prevalent. A number of major GIS meetings around the country are attracting an increasing number of lawyers. I doubt this is a good sign. The concept of recouping fees on data is obviously attractive, but there are a number of difficulties. The characteristics of the information I mentioned earlier contribute to the challenges of managing and charging for digital data. Data are hard to police and hard to track administratively. Services may be easier to charge for, and specifically attach a value to, than actual data sets.

Several other issues with geospatial data among Federal agencies are being recognized, one is that data are often out of date. This does not mean that older data are not valuable--certainly they are for trends--but often we want more current information and many parts of the country are not available in digital form at needed resolutions. Unknown quality is another issue. Often we find that when we share data, we do not know what we have: we get a data set from someone else and you do not know when it was last updated, what was actually delineated, or whether, for example, in a hydrologic data set every stream was delineated or only the perennial streams. Obviously, this creates challenges in making use of the data.

Another issue we are only just beginning to understand is that we spend a phenomenal amount of money to collect data, but very little to maintain or manage those data over time. We have not done a very good job at educating executives that beyond the need to collect data there are the requirements to manage them and keep them current. Funding must be allocated for data maintenance and management just as it is for hardware and software. Finally, data are incomplete; we do not have comprehensive digital geo spatial data at the resolutions we need for many places.

Ideally, we would like to evolve to a data management model that provides in any given space or ecosystem or place a means to establish common linkages of data sets that are being collected and ways to integrate them. This is not necessarily a centralized data base. It is very much a distributed activity, but we need someway to link the data that are being collected by different agencies at a given location. Ideally, we should not just have a data base but actually be able to look at the implications of policy recommendations representing different views based on the integration of that information. I have been struggling with how to think about some of this, and I believe conceptualizing data bases based in place rather than layers may help. Ecosystems are places, and we must begin thinking about managing data, and perhaps collecting it differently than we used to.

A couple of data models are often used when talking about GIS. One is to think of space as being empty, but with things in it in different places. The other is to think in terms of layers, that there is something everywhere for soils, for vegetation, and for hydrology. In reality there may be many things in certain places; there may be fewer things in other places. As we conceptualize organizing information related to place, this should affect how we manipulate geographic data. Many GISs function based on layers of information that are overlaid, assuming that there is somewhat of linear approach to the way things are actually distributed. There are fewer that actually function based on the unique characteristics and relationships in a place, where you may not have the same information everywhere. This is more of an object-oriented approach, which establishes linkages between things in space so that rare species or unique occurrences can be taken into account.

The advantage of looking at things this way is that decisions about how we manage resources aren't made based on consistent layers; they are made on tradeoffs based on what actually exists in the geographic area. Unusual events can be easier to acknowledge when thinking about information organized this way. And the notion of organizing by place or landscape fits better with various ecosystem models. If you think of a guild or community in an ecosystem where there are relationships between feeding habitats and how individuals forage for food, and relationships between individual species, it is easier in some ways to conceptualize a series of relationships between vegetation and fauna, than to sort through layers of data that must be overlaid. I am not an expert in the software approaches for this model, but I am trying to conceptualize the approaches we need for arranging information to answer the questions we face.

What we do in the layer-based system is overlay. That's a very common GIS approach. What we might do in a place-based system would be something more akin to forming a region based on characteristics of geography. Linkages between the characteristics of things that inhabit that space would be established. Rather than necessarily resulting in a map, statistics and characteristics of those regions that were formed would be generated.

Another advantage of thinking about organizing information in terms of places is that we might better describe what is unique about that place. One of the things I have always struggled with in the use of GIS is when I look at the screen and see the various layers of rivers, and roads, and so forth--it always seems so artificial to use a cartoon representation. When I am actually walking around in that space, it is a combination of variables that makes it unique, special, and valued. It is difficult to capture that uniqueness looking at a layer of roads and a layer of rivers. And yet, we are using our GIS to conduct real analyses--to make decisions about trade-offs, the locations of businesses or toxic dumps. This picture of reality seems so artificial. Somehow I think place-based organization of data gives us an opportunity to be more definitive about what makes a place more unique, and how it fits in with other places; how it contributes to the overall value of the region; how it is representative of other places; and how it might be managed.

The tools that will contribute to being able to think about place-oriented information management are some of the ones that you're familiar with--they may be part of your operating environment now. We recognize that there are numerous institutions that are collecting data, so a means to communicate is important. Internet is an informal means of grouping individuals of like interests to exchange ideas in a "virtual community"; but certainly we are also seeing people coming together in a proliferation of watershed management councils, and bioregional activities. I think one of the first ecosystem-related workshops I encountered was on the Sierra ecoregion. We are seeing these proliferate around the country as a means for institutions and individuals to come together to think about a place, to think about the common characteristics and requirements, and to communicate about it. The concept of virtual communities on the "Net" is an interesting one, but we must consider the implications of separating community from place. Internet offers significant means to communicate about the availability of data. As a means to search for and acquire data, WAIS (Wide Area Information Servers) and Mosaic are two of the most popular tools in use now. It is also reasonable to assume that they will continue to improve with time.

New means to document and understand data quality must be found if we are going to share data, and sort and sift through all that is available. We have done quite a bit of work on something called a metadata standard. Means to consistently collect data must be improved if sharing is to be facilitated. We are increasingly moving into a world where that stovepipe approach--where data you have collected are only useful to your organization--has become a luxury. We are living in a world of limited resources, no one can collect all of the data they would like to have. We need to interact with the other likely users of the data we collect to know what their requirements may be, and put into place standards and guidelines. This is true at all levels of government. State and local governments are willing to work with commonly developed guidelines and standards for how to collect data. If we can put these into place; not necessarily to generate identical data everywhere, but based on common needs among the players in a given geographic areas; we might actually save money and have better data. Finally, for ecosystem assessment we need tools that provide means to organize data by place which GIS certainly provides; means to analyze--again GIS--and various models; and means to display--visualization tools and GIS.

Data about a place or space are inherently spatial. This is why the National Spatial Data Infrastructure (NSDI) is important, because some of its activities, though not specific to ecosystems, are the underlying fabric of place-based approaches to arranging spatial data. I work with the Federal Geographic Data Committee (FGDC). This committee was established by the President's Office of Management and Budget (OMB) and is currently chaired by Secretary of the Interior Bruce Babbitt. It has policy representation from all the major Federal agencies that have anything to do with spatial data. The major goals that we have been working on relate to facilitating access to and use of geospatial data for a variety of purposes. This means, facilitating finding of data that already exist; facilitating means to understand what data you have when you do get data from someone; and also understanding how to better integrate some of these data. We would like to see all of this done more cost effectively, rather than every ecosystem analyst digitizing the same 7-1/2 minute quadrangle.

I would like now to summarize highlights of the Executive Order (EO #12906, April 11, 1994). By January 1995, Federal agencies are to document new geospatial data with the metadata standard; the FGDC is to develop a document outlining partnership strategies; and a plan for a framework data set is to be developed. If any of the information you collect falls within the definition of geospatial data, then you are required to use the metadata standard to describe it, and you will serve that metadata to the network. This activity is referred to as the National Spatial Data Clearinghouse. Within a year of the Executive Order, by April 1995, Federal agencies are to develop plans and schedules for looking at the data already held; specifically addressing how to make it accessible to the public. Agencies will also think about adopting procedures for using this Clearinghouse to search for data before expending Federal funds for the collection of new data. The Clearinghouse is intended to involve all levels of government and a variety of institutions, public and private, and academic. It is an ambitious activity. We have just initiated a competitive agreements program to provide small financial incentives to State and local governments to establish "nodes" that will describe what data exist in their geographic area.

We have been working to adapt some of the Internet tools, such as WAIS and Mosaic, so that data can be searched for spatially. This will contribute to the ability to assess the availability of data by place. Specific coordinates or a geographic area can be defined within the network searching criteria to pose the question of what data exist. This requires that within the metadata standard, the "footprint" or geographic coverage of various data sets be defined. Given that the President has signed an Executive Order that says that federal agencies will tackle this, we are optimistic that there is progress in thinking spatially.

I mentioned that data vary by geography based on applications. We often do not design our national programs with this thought in mind. For example, the National Mapping Division of the U.S. Geological Survey has mapped the country at consistent scales. There is a 1:24,000 scale base in paper form for the entire nation that was completed about 2 years ago. It took about 100 years to complete. Actually that's not quite true. They really didn't start this series until after World War II. This would be an extremely valuable base if it was digital and current everywhere, but it is not. At current levels of funding within the USGS, it will be approximately the year 2040 before it is completely digitized. Most people think that is too long. Because it is not officially available, there is a lot of redundant digitizing going on. I often talk about students digitizing in closets, automating these maps because they are not available as a standard base.

Recently the FGDC has been exploring the concept of building a digital framework data set. There are several data sets that are critical to almost everybody using a GIS. These act as the base for additional data collection, and include roads, rivers, terrain, administrative boundaries, and elevation. Additionally these must be accurately registered to the Earth through geodetic control, and many people also want an image base. These seven "themes" are conceived of as comprising the framework data set. OMB is considering the value of creating a national framework that would have the foundation data sets required by most users, so that duplication could be minimized. This would be a shift in thinking about how we manage information and how we work with partners that might contribute spatial data in any given place.

The requirements for these data may vary quite a bit in resolution. In urban areas, we find that street centerline, for example, may be needed at a very accurate resolution, whereas in more rural areas of Montana or Alaska, one meter resolution is not necessary. The idea of variation in resolution is one characteristic of the framework that is being discussed, as well as the possibility of accelerating the digital creation of these themes of data that would form the foundation for other data collection activities.

These institutional discussions will help to put a plan into place by January 1995, as called for in the Executive Order, which could have a major effect on the activities of Federal agencies involved in geospatial data collection. The expectation is that numerous institutions would contribute to development of these data including local government. Many counties are creating street centerline files that would form part of a national database. We see the national spatial data infrastructure as content in the national information infrastructure. Additionally, NBS has conceived a national biological information infrastructure with a biodiversity clearinghouse. All of these might be thought of as tools for managing and sharing information for ecosystem management.

DISCUSSION FOLLOWING NANCY TOSTA'S PRESENTATION

Question: It is interesting how you have contrasted the concept of place to layers as the future direction of GIS function. I know current GIS technology and software is based on a layering of information. What would be different about the new GIS technology that needs to analyze data coherently on a spatial basis? Can you also use current GIS technology for that purpose?

Response: I think you can use current GIS. It is more the algorithms of the models that are fundamental in the GIS. It is establishing linkages between the data elements, and as I said, a lot of the object oriented data models are really sort of place-based or they are built on the concept of relationships between things in space. For example, you could have a river--rather than having simply a hydrology layer, that would show relationships between terrain, the water course, location of the dams, and watershed boundaries. These might be linked and represented differently within the construct of the GIS than simply "here is a hydrology layer" and "here's an elevation layer," and they can be overlaid. There would be established linkages between the features. I know there are vendors working on this. So I would not give up on GIS. GIS is the tool we have to work with.

Question: You talked about various clearinghouses for data information. I think there is a lot of duplication of effort here. There is the National Aeronautics and Space Administration's (NASA's) Global Change Master Directory, for example, the National Biological Information Infrastructure, and EPA's Environmental Mapping and Assessment Program (EMAP). Is there effort to increase interdependence between these organizations so that they might all work together?

Response: Very much so. I did not specifically say it, but the clearinghouse is distributed. There is nothing centralized here. For example, the USGS Global Land Information System (GLIS), EROS Data Center's image management system or accesses to NASA's Master Directory would be linked as nodes in the National Spatial Data Clearinghouse network. You would find this inventory of data with unique tools for the Network, such as WAIS. NASA is a member of the Federal Geographic Data Committee, and they just recently agreed to take their DIF (data interchange format) and enhance it so that it meets the requirements of the metadata standard. The metadata standard is a little more robust than what you currently find in the DIF.

So there is definitely overlap, but there are also connections and as it sorts itself out you will find agencies with Mosaic home pages pointing to every other agency. I know if you use mosaic now, you often get lost in the Net. You say, I have been here before, but I did not come this way. We will see some of that sorting out. The tools that are going to be most valuable in the future are going to be those that allow us to search and filter the information that is out there on the Net. But I will admit there is some duplication now, although there is discussion going on between the global change community, for example, and the Federal Geographic Data Committee. We are all just learning how to do this. It is valuable to have different approaches. Also, when you look at an agency's culture you find different people sitting on different committees, and some of them talk to each other inside their organizations and some of them do not. So we are going to evolve at different rates inside the various organizations.

Question: I have a couple of questions. One is, as we move away from the layering systems, how does that affect costs? And the other more generally is how do you see the costs of data acquisition and presentation changing in the future?

Response: I'm not sure I can say. The evolution from layering systems is going to be a natural one as we demand more analysis capability. In some ways, it probably will be a cost we cannot avoid. I have no idea what that cost is likely to be. It could be that the shift to object-oriented analysis would simply an enhancement to whatever software you are using, and in the next release you would have the ability to be a little bit more sophisticated. But that's probably a naive statement because I suspect there will always be data conversion costs. We may have to rebuild our data sets to conduct these types of analyses. The costs of data acquisition are likely to go down because of the number of activities. For example, I did not say anything about small satellites, but this is a major remote sensing activity that will come on line within the next few years that will result in more data, at higher resolutions, than any of us ever knew we needed. Some of those costs now--the actual field collection costs--will probably stay the same as they always have, although if we have more specialists out there analyzing data, maybe those costs will go up. But the tools -- such as hand-held Global Positioning System (GPS) geographic locators and "Newtons" (portable fax equipment) -- will make data collection more efficient and perhaps more accurate. In the long run, we are going to have more data than we know what to do with--whether it is the right data will be one question, how we optimize how we manage it will be another.

Question: To what degree have agencies involved in the Federal Geographic Data Committee met with the management community at their agencies? It is one thing to discuss the philosophical relationship between all of these variables which, in fact, are not necessarily being communicated to those who count, namely Executive Branch managers.

Response: I assume you mean the natural research management community. What I talked about here in terms of this place-based notion is not a common discussion within the FGDC. The spatial data framework, the metadata standard, and the clearinghouse--all those are the FGDC initiatives. They are not exactly looking at how data are actually integrated in a GIS. They are facilitating access to data. They deal with data standards within the user community.

For example, the National Oceanic and Atmospheric Administration (NOM) in the Department of Commerce has responsibility for the bathymetry subcommittee of FGDC. So NOM is dealing with this notion of how do you represent shoreline. Is there one way to represent shoreline, and, if so, can all the off-shore resources be associated with it?

The OMB Circular (A-16) that created the FGDC, also charged various Federal agencies with coordinating thematic interests of all the other agencies that collect data related to that theme. Agencies are talking to data users to try and set standards. Every standard proposed by the FGDC goes out for very broad national review. They are announced in every GIS magazine, discussed at GIS conferences, and many of the specific subject area conferences--such as the hydrographic conference that occurred recently. So we try to involve as many viewpoints as possible in all of this. The only way you can have standards and guidelines accepted is if they come from the bottom up. My perception is that they seldom work if dictated from the top down. So we have tried to adopt the bottom up approach.

Question: Actually, I would like to make a comment while we were getting our technology demonstration together. That was a real good point to bring up. Are we collecting the right data? My thought is that it changes with topic. The right data for global change issues are very different from the right data for water quality issues, for example; and we as a society just have to be aware that our data needs change. I'm interested in rainfall which is one factor that determines water availability. Look at what's happened in the last 20 years in the Rio Puerco watershed in New Mexico. Is this possibly a harbinger of some of the things that are coming with global change? The right data is both what we collect for the problems we have to date and those we may foresee for the future.

Response: Right now, the efforts of the FGDC are focused on the minimal data sets that most people need--not all variations on themes. But you're right--data needs do change. Also, other Federal and institutional entities do consider the interdependency of efforts such as ecosystem management and global climate change research, and analyze how their respective needs for data might be similar, and where there may be overlaps and opportunities for resource sharing.

APPLICATIONS OF GPS AND GIS TECHNIQUES
IN MIGRATORY BIRD MANAGEMENT

Prepared by D. Alan Davenport (6)

Biologists in the Office of Migratory Bird Management have been conducting aerial surveys since 1955 to assess breeding waterfowl populations and habitat. A large percentage of the northern breeding range (figure 1) is sampled each year. Waterfowl are counted and annual population estimates are generated. These statistics are used to establish annual hunting regulations, for example.

One of the most important factors in conservation and wildlife management is the habitat and its condition. No organism can thrive under poor habitat conditions. Days could be spent discussing problems, progress, and other factors related to habitat studies. Global Positioning System (GPS) and Geographic Information System (GIS) techniques have the potential to bring the monitoring of wildlife populations and their associated habitat into closer focus for more detailed study and analysis.

How can population counts be more closely related to the habitat? This should be relatively simple in a small study area, but it is much more complex when dealing with thousands of square miles of very diverse ecological compositions. Most of the survey transects are divided into segments, usually 18 miles long. Survey data are recorded by transect and segment, so the information can be analyzed by segments. However, an 18 mile segment can contain many different types of habitat.

A discussion follows on the history of some of the efforts to develop techniques to link waterfowl habitats and population counts. In 1990, researchers at Patuxent Wildlife Research Center (now Patuxent Environmental Science Center) were using hand-held LORAN units to obtain latitude and longitude data for areas of potential study. A portable computer was used to process geographic location information generated by the LORAN receiver. Researchers then posed the question: Could a similar technique be used with the LORAN navigational units in a survey aircraft?

Aerial surveys are generally conducted by two individuals, the pilot and an observer. Each counts the birds seen on his or her side of the aircraft and records the observations on audio tape. To use the LORAN there had to be some automatic way to record the location since neither the pilot nor the observer would be able to look at the instrument panel and write down the latitude and longitude. A program for a laptop computer was developed, which when connected to the LORAN unit, would continually read the location information produced. Additional micro switches were rigged with the "push to talk" switches used by the pilot and observer. Closing the switch triggered the computer to record a location and time for that individual. This procedure enabled collection of reasonably accurate locational data (figure 2). The major problem, however, was that considerable time coordination was necessary to later be able to match the taped observations with the recorded locations.

An experimental flight was made in December, 1991, counting waterfowl encountered along a portion of the Chesapeake Bay. A basemap was created by combining a number of digital line graph maps of USGS 7.5 minute quad sheets, and the observation points and basemap were plotted using Arc/Info GIS software. The resulting plot (figure 3) indicated that the LORAN method of recording locations was one workable solution to part of the problem. There was still the habitat issue, however, and how appropriate and usable habitat information could be obtained. The ideal solution would be to use aerial photography, satellite imagery, or perhaps some other remotely sensed information; however, for the level and scale of these experimental observations, that would be too expensive.

Some time after the development of the LORAN system, the office procured GPS hardware and software. The basic difference is that with GPS, location is determined by triangulation from a number of satellites positioned in stationary orbits. This system is operational worldwide. LORAN calculations, on the other hand, are based on signals from a network of communication towers on the ground, and some areas of the world are not covered. GPS, therefore, appears to be the logical navigation choice of the future .

An initial GPS application was an attempt to record habitat along Breeding Bird Survey (BBS) routes. The BBS is a permanent roadside survey that has been conducted annually by volunteers for the last 27 years. There are over 3000 randomly established routes in the US and Canada and approximately 2500 surveys are conducted annually. BBS is now being performed under the National Biological Survey.

The GPS allows us to record points, lines, and areas. The points could represent the bird observations, the lines could represent roads, and the areas could represent habitat blocks along each side of the road. Points and lines were no problem to collect, but recording an area requires that points be collected along its perimeter. This proved difficult if not impossible, especially in trees where transmission of the satellite signals could not penetrate.

The possibility of collecting habitat data from the air was considered using one of the aircraft used for waterfowl surveys. A test route was selected and a flight was made along the route in both directions. The observer recorded habitat out to 200 meters along the right side of the road.

The GPS has a wand for reading barcodes referenced to a dictionary in the Datalogger unit. This feature permits rapid data entry. A number of broad habitat classifications were set up as line features and a barcode sheet was prepared (figure 4).

Each "line" of habitat was mathematically converted to a polygon up to 200 meters wide. They were initially located according to the position of the plane which was usually on the opposite side of the road. It was necessary to adjust polygons relative to the center of the road (figure 5). The habitats recorded on the test route were ultimately displayed by ARC/INFO GIS software (figure 6).

Using GPS techniques for aerial population monitoring surveys has some distinct advantages over the LORAN method discussed earlier. The GPS signals are available essentially anywhere in the world, the barcode scanner permits rapid data entry, and the software provided with the system automatically prepares data for analysis by a GIS. The main disadvantage is that another person is required to enter the data (figure 7). The pilot and observer operate as before, but now the voice record is a backup. The recorder listens to the comments of the pilot and observer and enters the sightings into the datalogger with the barcode wand. The man-hours associated with the extra person on the survey are minimal when compared to the hours that were necessary with older methods used to link the audio observations to the location points, however.

The GPS equipment was tested on several waterfowl population surveys in 1993 and early 1994. Survey techniques were improved with each trial of the system. Counts are recorded using a set of 36 barcodes for 1's, 10's, 100's, and 1000's. The GPS considers the entries as modifiers to the observation point which represents a species of waterfowl or whatever. Any number can be quickly entered into the system using the appropriate combination. A program was written to add them up before they get into the GIS.

The Trimble GPS unit used comes with software to process data from the datalogger. Data can be displayed on the screen, queried, edited, measured, and plotted. Each point is represented along the flight line by a cross. Any point can be queried to display the attributes that were recorded with it (figure 8). After the data are incorporated into the GIS, attributes can be displayed and queried using ArcView software. In one example, only flocks of over 500 birds were displayed along with an identification number assigned by the GIS (figure 9).

At the beginning of this year a request was made to investigate the possibilities of using GIS and Breeding Bird Survey data to explore possible relationships between the Conservation Reserve Program (CRP) lands and abundance of birds. Data were acquired for the first 11 CRP sign-ups and a quick tally disclosed that 92.7 percent of the acres were in grassland or wildlife habitat.

A framework for measurement was established by arbitrarily splitting the area into five density categories based on the number of grassland CRP acres per county: [0] No CRP acres in the county; [1] < 0.75 percent CRP (50 percent of the counties with CRP); [2] 0.75 percent - 2.35 percent CRP (20 percent of the counties with CRP); [3] 2.35 percent -7.5 percent CRP (20 percent of the counties with CRP); [4] > 7.5 percent CRP (10 percent of the counties with CRP).

Distributions of several species of grassland birds were examined, one of which is the Grasshopper Sparrow (Ammodramus savannarum) (figure 10). Population change was estimated and patterns of change were plotted. For purposes of a simple illustration of this concept of using GIS analysis, an area was selected which was bounded on the north by 45 degrees N. Latitude, south by 40 degrees, east by 90 degrees W. Longitude and west by 95 degrees. This GIS analysis was done by overlaying the CRP data with the bird distribution data and evaluating coinciding areas.

For example, the GIS can show and compute the size of all the areas where birds were increasing in the greatest CRP density (figure 11). In this case, the GIS reports 6,204,258.8 hectares for the CRP of which 4,376,924.7 hectares or about 70 percent had increasing Grasshopper Sparrow populations.

Results of this nature are interesting, but must be interpreted with caution, however, because of other factors which can confound the underlying trend analyses used with the BBS data. Factors such as drought, conditions on the wintering grounds of migratory birds, sample sizes in the surveys, etc., are not reflected here. Also, the distribution of the BBS survey routes is less concentrated in the prairie areas where the greater amounts of CRP lands are found. The value of the analysis, however, is that it quantifies apparent relationships between the birds and their habitats that can then be scrutinized more closely to determine the significant relationships.

In conclusion there were several policy-relevant questions that were to be addressed as they pertain to the material presented in this paper:

• What are the opportunities and limitations of these tools;

• Where is the development of these tools headed, and how rapidly are they changing;

• What are the costs associated with the tools, and how do those costs compare with the cost of data acquisition;

• Is there a need for new or different data based upon the capabilities and opportunities of the tools; and

• What is the potential use of these tools for informing the public decision making process and contributing to other national goals?

GPS equipment is limited only by its ability to receive signals from enough satellites (at least two, preferably three or more) to compute a location. GPS units are being made smaller, durable, and more reliable. GIS software is available from a number of vendors at costs commensurate with versatility and complexity. Data used in a GIS must be geo-referenced to be used, therefore older data sets may be of little value unless they can be re-worked. A number of commercial companies are involved in marketing geo-referenced satellite imagery and data from other sources at varying costs. The use of GIS techniques is rapidly expanding worldwide.

Endnotes

4. Wayne A. Morrissey is a Senior Research Assistant in the Science Policy Research Division of the Congressional Research Service.

5. Nancy Tosta is Chief of Geographic Data Coordination for the National Mapping Division of the U.S. Geological Survey. This paper has been edited from a transcript of the original presentation.

6. D. Alan Davenport is a GIS Specialist at the U.S. Fish and Wildlife Service. This paper has been edited from the transcript of the original presentation.