Dr. Dot Helm, University of Alaska Fairbanks, Palmer Research Station, was brought on via Cooperative Agreement to serve as Principal Investigator for the vegetation component. Dot’s first assignment was to prepare an evaluation of the current protocol/data, with recommendations. She was also asked to evaluate the usefulness of previously studied vegetation plots and satellite imagery options; to ferret out the needs of other investigators for vegetation data, and scope out preliminary conceptual models of vegetation change in the park. Carl Roland, who had been hired by the park to collect the vegetation data during the summer of 1998, was kept on over the winter with USGS funds to provide continuity and to analyze all the vegetation data collected thus far by the monitoring program. This analysis was critical to our effort to evaluate the current protocol and its value for the future.
- Roland, Carl. January 1999. Summary
and analysis or vegetation data from Denali long-term ecological monitoring program
permanent plots.
- Helm, Dot and Carl Roland. February 1999. Evaluation of existing Vegetation Protocols. Denali Long-term Ecological Monitoring.
Carl’s report included an analysis of weather for the 1992-1998 period, based on the Doggy Data (Headquarters National Weather Service site). This is a valuable portion of the report that should be of interest to all investigators. This section of the report is excerpted here:
- Temperature, precipitation and selected climate indices for the period 1992-1998.
In early March, Karen Oakley, Dot Helm, Carl Roland and Page Spencer met to review Carl’s and Dot’s findings, and discuss and define a new set of objectives for the vegetation monitoring component. The results of this meeting were captured in a memo to Gordon Olson. The purpose of the memo was to describe how our thinking has evolved and as a vehicle for soliciting input about the future of LTEM at Denali.
- Memo to Gordon Olson re: Recommended changes for the vegetation component of the Denali LTEM program
Based on our analysis of the past data and discussions of new objectives for vegetation monitoring, Dot prepared a study plan laying out work to be conducted in the next two years to complete the vegetation monitoring protocol. This plan proposes a new design and suggests a variety of efforts that could be carried out to further develop and test the design.
Dot, Carl and Karen met in the park on June 14-15, 1999, to discuss priorities for 1999 field work. Dot prepared a more specific study plan to describe the work. The objectives for the summer are to try out three different methods of quantifying plant cover and species richness: (1) point-intercept transects, (2) Modified Whittaker plots, and (3) 1999 pilot version of Plant Structure component of the USFS Forest Health Monitoring program. The techniques will be compared in forest, shrub, herbaceous habitats. In addition, Dot will experiment with using digital photographs to get an index of berry productivity.
Four other items of interest with respect to vegetation:
- Dot Helm Attends USFS Forest Health Monitoring Training in Oregon
In late May, Dot Helm was given the opportunity by the
USFS to attend a portion of their annual training for the crews carrying out the Forest
Health Monitoring program in the western states (Washington, Oregon, California). The
Forest Health Monitoring program is of interest because it uses a grid design, collects a
variety of data that may be of the type that could meet some of Denali’s objectives,
and will presumably be coming to Denali park anyway. Being able to attend the training
provided Dot with a chance to become better acquainted with FHM, which will help our
evaluation of its appropriateness for Denali.
- Tree-ring Study in Rock Creek
A Cooperative Agreement with Glenn Juday and the Tree-ring Laboratory at the University of Alaska Fairbanks was set up in 1998 to conduct a small study of white spruce in Rock Creek. Trees were cored in September 1998 in the vicinity of the LTEM plots established by Roseann Densmore. Weather data from the National Weather Service station at Headquarters (the doggy data) were used to analyze tree growth in comparison to climate. The data on the spruce trees cored near the Treeline plots have been used in an analysis of white spruce treeline across Alaska.
The major finding of the tree-ring work at the Rock Creek reeline site is that recruitment of white spruce at that site has occurred since the end of the Little Ice Age (1830-1870). The continuous recruitment and lack of trees recruited before 1830 is consistent with a recent (since the Little Ice Age) and steady treeline advance in this part of the Alaska Range. They also found that treeline trees at Denali are generally vigorous (that is, they are not under moisture stress as has been found some other locations such as Fairbanks), and have been reproducing steadily or at intervals (related to big cone crop years) in this century. Big cone crop years have included 1912, 1915, 1926 and 1940--all years with favorable spring weather. Many of the trees at treeline currently date from the 1940-41 seed crops.
For the complete story on treeline in Rock Creek and elsewhere in Alaska, please see the paper published in the proceedings of the conference "Sustainable Development in Northern Timberline Forests" held May 10-11, 1998,Whitehorse, Canada. The proceedings were edited by Sakari Kankaanpaa, Tapani Tasanen and Marja-Liisa Sutinen and published by the Finnish Forest Research Institute, Ministry of the Environment of Finland.
- Juday et al. 1998. Recent dynamics of white spruce treeline forests across Alaska in relation to climate.
Other important findings from the tree-ring study come from the analysis of the Doggy Data. Glen found a significant increase in temperature after 1967 that did not jive with temperatures in Fairbanks (in other words, Denali did not experience its own instantaneous climate change in 1967). He found that the location of the sensor had been moved in that year, leading to a consistent change in the temperature being measured. Anyone attempting to use the Doggy Data as a series that crosses the year 1967 needs to be aware of this situation. Glenn has suggested that we need to develop a "standard, adjusted" Doggy Data set, so that everyone is using the same data for long-term analyses.
The other finding is that the 1980s and 1990s have been very, very warm (compared to the past).
- Bumper Spruce Cone Crop of 1998
Once it became obvious that 1998 was going to be a bumper year for spruce cones, Carl Roland and Page Spencer initiated a study to quantify the amount and viability of spruce seed produced by trees in eight separate sites along the road corridor of Denali NP/P. The central question this study was designed to answer is whether the bumper cone crop will actually result in a large pulse of viable spruce seed being dispersed across the landscape. In addition, this study provides useful data on the mean and variance in seed production among individual cones on a tree, trees within a site, and sites on the landscape. These data will be useful for modeling spruce dynamics in the future. Incidentally, we documented the insect fauna that parasitizes spruce cones in these sites.
- Plant Inventory Work in 1998, Continues in 1999
Another important development in the area of vegetation at Denali was the inventory work carried out by Carl Roland in 1998. Carl did an analysis to prior to the field work to identify gaps in understanding and to target locations for site visits. A total of 25 such sites were visited. The results of this field work included the collection of more than five hundred voucher specimens of vascular plants that documented the occurrence of six families, 20 genera, and 78 species not previously known from the Park. A majority of these new records were from sites on the south side of the Alaska Range, where relatively little plant collecting has occurred in the Park. Carl successfully requested additional monies from the NPS national inventory program to continue inventory work in the summer of 1999.
- Roland, Carl. 1998. Report on Floristic Inventory Project. Denali National Park and Preserve. Summer 1998.
NOTE: This work is not funded by USGS but I have included it in the synthesis because the findings are important to our growing understanding of Denali vegetation and how it should be monitored.
Aquatic Invertebrates and Stream Communities
A Cooperative Agreement was developed with Dr. Sandy Milner and Dr. Sarah Conn, University of Alaska Fairbanks, to continue work related to monitoring of stream biota. The first task in this agreement was to prepare a summary of the work conducted since the beginning of the program and a study plan for the next phase of work. An important goal in summarizing the past work was to place the Denali work in context with other efforts to monitor stream ecosystems, in Alaska, and worldwide.
The work carried out by Sarah Conn at Denali, 1994-1996 for her Master’s thesis and doctoral dissertation focused on classification of Denali streams based on their aquatic invertebrate populations. Six stream types, ranging from clearwater streams to glacial rivers, were defined. Because of its mountainous location, Denali streams are primarily headwater streams. In comparison to streams at lower latitudes, Denali streams have a low diversity of invertebrates. The majority of individuals are chironomids (non-biting midges).
Stream invertebrate populations are commonly monitored throughout the world as indicators of stream health. However, two approaches have emerged, and there are a number of related sampling issues that are still being hotly debated amongst aquatic ecologists and the regulatory agencies that have supported the development of stream monitoring protocols. (These issues include such things as whether sampling should be quantitative or qualitative, what parts of a stream should be sampled, the type of sampling device, and how many organisms in a sample should be counted.) One approach, developed in the United States and advocated by the U.S. Environmental Protection Agency and many states, focuses on the use of metrics and comparison of the biota in a sampled stream to a pre-defined reference condition. The other approach, developed in Great Britain and now being used in other parts of Europe, Scandinavia, and in New Zealand and Australia, focuses on multivariate analysis and prediction of what taxa should be present, based on physical and chemical characteristics of a stream.
In the Denali work, both approaches are being evaluated (i.e., they are not viewed as necessarily incompatible). Two observations, however, are leading us to suspect that the multimetric approach may not be useful at Denali. Principally, as there are so few taxa, many of the commonly used metrics are not suitable and show a high degree of redundancy. The metric scores were highly variable within a reference group which greatly reduces the ability of the metric to detect subtle impairments of the systems. In comparison to metrics for streams in lower latitudes, the metric scores for Denali streams suggest that these streams are highly polluted or degraded, yet we know these are pristine streams. Another observation from the Denali data is that neighboring streams may support very different invertebrate communities. Defining a "reference condition" for the general Denali ecoregion, rather than basing it on other factors such as the physical characteristics, or water chemistry, could lead to huge variability within the reference metrics, and hence be very inaccurate.
The Cooperative Agreement for the current work was put together very late in 1998, so a full field season was not possible. One late season trip was made to help continue the data series in the streams along the park road that been sampled in prior years. For a complete description, see the trip report.
The focus of the work in the winter was for Sarah to complete and defend her dissertation (which she did), and then write what we called a "variables assessment" report and study plan for the next phase of work. In the end, we merged the two documents into a single report which summarizes the work done in 1994-1996 and explains it significance and places it in context of bioassessment work in the U.S. and worldwide, and lays out the plans for the future.
In summary, the objectives of the continuing work are as follows:
- Compare the predictive (multivariate) and multimetric
bioassessment techniques sensitivity in their ability to detect impaired sites.
- Compare Surber samples with other methods,
particularly the qualitative techniques and sub-sampling methods used in the Alaska Stream
Condition Index (Major et al. 1998). For each method the minimum number of replicates
required to provide accurate results would be determined.
- Examine the biovolume of collected
invertebrates at sites in different groups and from impaired areas to examine the
potential of biovolume as a bioassessment tool for Denali NP.
- Using the new EPA-RBP protocols, examine the periphyton
communities and explore their role as possible bioassessment tools both in terms of
biomass and community composition.
- Try to document the range of natural variation
and identify some of the causes of the large year-to-year variation observed in the
macroinvertebrate communities of Denali NP (e.g. climatic effects etc.).
- Examine the overall physical stability of
the systems and explore links with community persistence, and inter / intra annual
community variability.
- Examine food availability and retention of
the systems in more detail to develop a great understanding of links between invertebrate
communities and the riparian zone.
- Chironomid identification - what taxa are
present, do they vary between classification groups, do they show such a high degree of
inter / intra annual variation.
- Effectively communicate the results of the research to park staff and the wider scientific community.
 |
a tour of Denali Stream Classifications |
Small Mammal Population Monitoring
Monitoring of small mammal populations in the Rock Creek drainage has been a part of the Denali LTEM program, under the leadership of Dr. Eric Rexstad (UAF), since its beginning in 1992. The method used employs live-trapping and is intensive: it involves several trapping occasions over the snow-free season and the mark-recapture (using passive integrated transponders-PIT tags) of individuals. The data produce point-in-time population estimates as well as estimates of survivorship and recruitment. The work at Denali is relatively unique: the more common method used by both researchers and resource management agencies to track small mammal populations in Alaska has been snap trapping, which only generates a population index.
Some of the questions that have arisen after the first 6 years of small mammal monitoring in Rock Creek include: Is the spatial scale of the monitoring effort appropriate? How could the effort be expanded to provide indicators of small mammal population status and trends for more of the park? Do we need a population estimate, or could we live with a population index? What are we learning about the ecology of small mammal populations? What is the value of the data to park management?
A new Research Work Order was developed with Eric Rexstad to:
- continue the small mammal population monitoring according
to the original design (with NPS funding),
- increase the effort put into data analysis, and
- grapple with the questions about the overall small mammal monitoring component of the LTEM program.
Another component of the Research Work Order was to address integration of monitoring data, described in the following section. To tackle both assignments (small mammal issues and integration), Ed Debevec was brought on full-time for the two-year period to work with Eric on this effort.
"Scaling Up"
One of the first assignments concerned how the spatial scale of small mammal monitoring could be expanded. Ed and Eric prepared a "thought piece" entitled:
In this report, they tackle the question of population indices versus population censuses and estimation and questions of temporal and spatial scales of sampling. As we all have found, it is difficult to impossible to answer some of the questions for just one component of the monitoring program. If we are going to have a coherent and integrated monitoring program, we need to nail down the goals and come up with a design that matches those goals. As a vehicle for getting some discussion going, they lay out a possible monitoring framework for the entire program.
The proposed monitoring framework begins with two assumptions: (1) monitoring should be conducted along the road corridor only and (2) monitoring can be conducted within sight of the park road. They propose a two-stage sample design, using 12 monitoring sites, and four strata, defined by major drainages that cross the park road. Monitoring sites would be randomly selected within each strata (within 5 km of the road), and subplots would be established for the various components (small mammal, vegetation, avian point counts, etc.) They also propose a rotating monitoring intensity, so that in any given year, some sites would be intensively monitoring, others would receive a non-intensive effort, and others no monitoring at all. One possible scenario would be that each site is monitored every 2 out of 3 years for a total of 8 visits over 12 years. This scheme would maximize our ability to make comparisons among sites.
Power and cost issues
A second "thought piece" was developed to discuss issues of power and cost as they relate to small mammal monitoring.
This report provides a general introduction to power issues, but makes an implicit point that is not usually emphasized about power: Power is only relevant for a clearly stated statistical test or comparison. Power has no relevance beyond the test and scenario for which it is calculated. We may have adequate power to detect a change we think it is important to detect, but may miss the change that actually occurs. Eric and Ed point out that we have a bit of a conundrum here:
" . . . On the one hand, we are asked to design a monitoring protocol that is all-encompassing and not tied to any foreseeable change, while on the other hand we must consider very specific changes in order to talk about power. . . ."
This conundrum brings up the overarching issue of what kind of monitoring program is this anyway? In monitoring that is used to direct very specific management actions, the question of power (and its required components--the effect size and direction you care about, and your tolerance for risk--your willingness to make a Type II error and the consequences of a Type II error) must be addressed. For monitoring that is more general, that is not tied to a specific management action, what are the important questions to answer relative to power?
Ed and Eric continue: Any consideration of power and small mammal monitoring must recognize the irruptive nature of small mammal populations. Questions about differences among years are moot: the answer is an unequivocal--"Yes, there are differences among years." The questions then become related to space: Are there spatial patterns in abundance or survival? This is a question that makes sense and provides a context for a discussion of power. To address the question of power to detect differences in small mammal populations among locations, simulations were used. Two scenarios were considered, using the currently used small mammal monitoring methods (12 secondary events over 4 days) and range of estimates generated from Rock Creek populations:
- Do we have adequate power to detect differences among two
locations?
- Do we have adequate power to detect trends among multiple locations (say along a transect)?
The results of the simulations suggest that power to detect a difference between two locations is reasonable. Similarly, power to detect a linear trend along a transect is also reasonable. This exercise gives us a general sense that we will be able to detect some changes. However, the main point to come out of the exercise relates to the question of what changes do we want to detect? Until we know that, power calculations are somewhat pointless.
So, what are we learning about small mammal populations anyway?
One of the important goals for this phase of work was to actually try to do something with monitoring data. To jump start this process, Oakley and Rexstad (foolishly?) committed to presenting a paper at a forest monitoring conference to be held in Guadalajara, Mexico, in early November 1998. Ed came on board just in time to play a major role in this effort. The paper presented an exploratory analysis of the small mammal and passerine bird data from Rock Creek. This analysis revealed few measures that were significantly correlated with small mammal abundance. A power analysis revealed that several more years of data would be required to determine if the lack of relationships was truly non-significant. This analysis demonstrated one of the challenges to using monitoring data to try to understand cause-effect relationships--a very long time series of data is usually required, and maintaining enthusiasm and support for continued data collection over such a long period of time can be difficult. For the full story, see the manuscript of the proceedings paper expected to be published in fall 1999.
- Abstract
- Manuscript
Over the winter, further analyses of the small mammal and weather data were conducted, leading to the development of a vole population model. This model supports the hypothesis that vole abundance in Denali is driven by environmental conditions with little carry-over effect from the previous year. The model will be presented at the Ecological Society of America annual meetings this August in Spokane, Washington.
General Monitoring Program Development Issues
Because the Denali LTEM program is being developed as a prototype, one of the goals of this phase of work is to capture lessons learned about monitoring program development. Karen Oakley (USGS--Alaska Biological Science Center) is working on this topic, as time allows. One aspect of this work has involved extensive review of the monitoring literature. In the course of this work, a Pro-Cite bibliographic database is being compiled which will eventually be served on this website.
Framework for Understanding Long-term Monitoring and Research Programs
A framework for understanding long-term monitoring and research programs has been developed. This framework recognizes three relatively distinct classes of activity: (1) long-term ecological research (sensu Likens ), (2) long-term ecological monitoring (sensu EMAP), and (3) adaptive management monitoring. The distinctions among these classes are still being considered, but so far the framework seems to provide a useful paradigm which could help ensure that when we talk about "monitoring," we are all talking about the same thing. The classes do differ in their purposes, how they would be set up, and what can be expected from them, so being aware of the differences is critical in any discussion of monitoring program development.
Monitoring Program Development Process
Flow charts showing the steps to developing a monitoring program are being evaluated and refined to incorporate information gleaned from the recent literature on monitoring. An important part of this work is to develop more explicit guidance for the parts of the process that represent trouble spots. Definition of objectives seems to be one of those trouble spots. Methods to set objectives and ways to ensure a logical connection between goals and measurable objectives are being developed.
A description and analysis of the early history of the Denali LTEM program was presented as part of the paper prepared by Oakley, Rexstad and Debevec at the forest health monitoring symposium in Guadalajara, Mexico. Two lessons about monitoring program development emerged. First, that it is very important to define the roles of everyone involved and develop good lines of communication. Second, it is important to not be in too big a hurry. The Denali experience reinforced the importance of clearly envisioning intended data uses before any data are collected and carefully matching the design to the objectives.
- Abstract
- Manuscript
Scenario Models
As a follow up to the Starfield modeling workshop held in October 1998, we are hoping to look at the use of scenario modeling and decision analysis as tools in objective setting and balancing. Scenarios are widely used in business and in the military, but their application to resource management questions has not been widely explored. We hope to develop several alternative scenarios for 2050 as a way to focus our attention on the most important changes the park would like to detect with its monitoring program.
Integration and Synthesis of Monitoring Data
Part of the new Research Work Order developed with Dr. Eric Rexstad, University of Alaska Fairbanks, asked him to work on how data from the various components of the monitoring program could be integrated and synthesized. Integration was always understood to be an important desired feature in the Denali monitoring program, but little integration of data sets has occurred thus far. One purpose of data integration is to use the monitoring data as fully as possible to improve our understanding of the ecosystem. The primary purpose is that data integration will eventually lead us to an assessment of ecosystem health. As mentioned above, Ed Debevec was brought on to carry forward the analysis of small mammal monitoring data and to begin our effort to integrate the monitoring data from all components of the program.
The first steps towards integration were taken with the small mammal and meteorological data from Rock Creek, as described above. Two additional projects are envisioned:
- A comparison of avian monitoring methods
currently used in the Denali LTEM program; and
- Measuring the Health of a National Park.
Comparison of Point Counts and MAPS
Two bird monitoring methods are used at Denali. Point counts are made which provide annual estimates of frequency of observation, as an index to population abundance. The Monitoring Avian Productivity and Survivorship (MAPS) program uses mist netting to catch and mark birds, thereby yielding estimates of productivity and survivorship, in addition to a frequency of observation. The point count method is relatively cheap and easy to implement; MAPS is intensive and requires considerably more effort. In this analysis, we will explore how the two programs overlap to see if the results agree and to what extent the methods are redundant.
Initial analysis of the bird monitoring questions has been completed and has been submitted for presentation at the upcoming AAAS meetings to be held at Denali Park in September 1999.
A Measure of Ecosystem Health
One of the most critical but neglected aspects of any monitoring program is reporting information in an immediately useful way. We will explore the use of summary metrics for reporting an annual measure of "health" for the Denali ecosystem. One approach we will explore will be to incorporate an indication of "standing crop" and "productivity" for each aspect of the ecosystem that is monitored. These metrics will give an idea of "how many" and "what direction" they are going. The metrics could be combined to yield a "gas gauge" or remain separate, as in an "instrument panel." The idea would be to explore options for reporting information from the monitoring program in a common way so as to promote a synthesis of the current state and trend of the ecosystem.
Lynne Caughlan, economist at the USGS-Midcontinent Ecological Science Center in Fort Collins, CO, is working on an analysis of cost issues in long-term monitoring. The first step in this project is a literature review. Many authors touch on the importance of cost, but nothing has been written that pulls all these snippets together. Once this literature review is completed, the work will focus on determining the true costs of the Denali monitoring efforts. The purpose of this work is to develop a realistic understanding of monitoring program costs such that the program is sustainable and cost-effective in the long-haul.
We have been dancing around the edges of the data management issue. Basically, I have tried to learn as much as I could about data management this winter. I wanted to be ready to interact with the Data Manager expected to be hired soon by the park. I also needed to develop guidance to the Principal Investigators being funded by the USGS as to what was expected of them. I clearly did not know enough about "data management" yet to do that. I am not sure if I am there yet, but some progress has been made.
Database Design
I started by taking several basic Microsoft Access (the NPS standard which will be used in Denali LTEM) classes, provided (free of charge) through the DOI Learning Center. I had prior experience with relational databases from working with Paradox, but had never had any formal training in database design. From these Access classes, I learned about the critical importance of database design (not for amateurs!). Good data base design requires extensive interaction with the data users, at the beginning. A clear understanding of how the data will be used is required. The tables and their relationships are sketched on paper BEFORE you even look at the computer. A lot of thought is given to the reports and forms that are desired. In designing databases, one should strive for at least 3rd normal form, and one should try to achieve 4th or 5th normal form if there are no compelling reasons not to.
The database design steps suggested in the Access class were as follows:
- Talk to users to find out what questions they will have of
the data and what kind of forms and reports they might need.
- Plan tables and relationships (on paper).
- Normalize the design.
- Create the tables.
- Create relationships.
- Load dummy data and test (run queries).
- Download real data, or Start data entry process.
- Create all queries, forms and reports.
- Automate, as needed (macros and modules).
- Distribute
While at the forestry monitoring conference in Mexico, I heard several papers relating to data management, one of which had some useful information related to design. Dr. Donald Henshaw, one of the data managers for the H.J. Andrews Experimental Forest Long-term Ecological Research (LTER) site, presented a very interesting paper on data management there. He presented a data management process. He emphasized that the database needs to be designed BEFORE any data are collected. Dummy data are used to test the database. Thus, significant interaction between the scientists and the Data Manager takes place before, rather than after, the fact. [I should mention that the LTER program in general, and the Andrews forest LTER site in particular, have really been the leaders in developing our understanding of data management and its importance in long-term studies, over the past decade or so.]
Naming Conventions
Another important tip I picked up at the Access classes was the importance of having some kind of naming conventions. Naming conventions are used to clearly identify the various objects (that is, the tables, forms, reports, and queries that comprise the database), and also to identify the type of data that goes into a field. Naming conventions are very important if other people are going to be using the database--clearly appropriate in the case of long-term monitoring. The use of naming conventions also seems clearly important in a case like Denali long-term monitoring where we are trying to build a program of integrated data. If everyone uses the same naming conventions, it will be a whole lot easier to make sense of someone else’s data files.
The naming conventions advocated by the people I was taking the class from (MicroAge) are the Leszynski Naming Conventions for Microsoft Access, which are proprietary, but available for a nominal fee ($10) from the Kwery Corporation (or call 1-800-AT KWERY). These conventions were developed by Stan Leszynski as a "global standard for VBA (Visual Basic) programmers." The nice thing about these standards is that they are already developed and ready to use, thus meaning we would not have to reinvent the wheel. We probably do not need to use the full Leszynski Naming Conventions which are aimed at full-time programmers. A Level One is described which is aimed at users whose work is with the database window and who do not develop applications. The Level One Leszynski Naming Conventions seem like a good place to start to get everyone in the Denali LTEM setting their databases up in a similar way.
A Preliminary Review of Some Denali LTEM Datasets from a Database Design Perspective
To get a better handle on our current state of data management within the Denali LTEM and what the next best steps might be, I let a small contract with MicroAge. Mike Robbins, an experienced Access programmer and consultant, reviewed several of the datasets as they currently are (small mammals, vegetation, birds, doggy data) and met with me to answer some of my questions. He then gave some rough-cut guidance on how we can improve our data management situation.
The upshot is that we still have a long way to go in the data management arena. It is unreasonable to expect each PI to become a database design expert, and we need to bring in someone with considerable database design experience to interact with investigators. It would probably help if everyone had a modicum of understanding about relational databases, to facilitate the interactions with the database expert.
Database Development for the Organ Pipe National Monument Long-term Monitoring Program
On my trip to Tuscon for the Status and Trends Program Review, I was able to meet with University of Arizona graduate student Susan Cherry. Susan is working on the Access database used in the Organ Pipe National Monument long-term monitoring program. She demonstrated the application she is building for them. My discussions with her reinforced the need for substantial interaction between the scientists and someone who understands databases in the process of database design.
Data Management Literature
I have tried to get my hands on as much of the published literature as I could find on the topic of "data management." Here’s what I have found so far:
I also found a great web site where the following publication can be viewed or downloaded:
This publication is really the latest and greatest and an excellent resource for anyone trying to learn about data management.
Another key reference in the data management and data integration area is:
- National Research Council. 1995. Finding the forest in the trees. The challenge of combining diverse environmental data. Selected Case Studies. National Academy Press, Washington D.C. 129 pp.
Date: July 2, 1999
This list catalogs presentations and contributions to the scientific literature that can be traced to the involvement (primarily funding) of the USGS (and formerly the National Biological Survey) in the development of the Denali Long-term Ecological Monitoring Program.
Published Papers
Juday, G.P., V.Barber, E.Berg, and D. Valentine. 1998. Recent dynamics of white spruce treeline forests across Alaska in relation to climate. Pp. 00-00 In Proceedings of the conference "Sustainable Development in Northern Timberline Forests" held May 10-11, 1998,Whitehorse, Canada. Sakari Kankaanpaa, Tapani Tasanen and Marja-Liisa Sutinen, editors. Finnish Forest Research Institute, Ministry of the Environment of Finland.
Paton, P.W. and T.H. Pogson. 1996. Relative abundance, migratory status, and habitat use patterns of birds breeding in Denali National Park, Alaska. Canadian Field-Naturalist 110(4):599-606.
Rexstad, E. 1994. Detecting differences in wildlife populations across time and space. Trans. 59th No. Am. Wildl. And Natur. Resourc. Conf. (1994) pp 219-228.
Thorsteinson, L. and D. Taylor. 1997. A watershed ecosystem approach to monitoring in Denali National Park and Preserve. J. American Water Resources Association 33(4):795-810.
Theses
Conn, S.C. 1998. Benthic macroinvertebrate communities in the rivers of Denali National Park and Preserve, Alaska: An approach for watershed classification and ecological monitoring. Ph.D. Dissertation, The University of Birmingham.
Furtsch, P.R. 1995. Techniques for monitoring density and correlates of interannual variation for northern red-backed voles (Clethrionomys rutilus) in Denali National Park and Preserve. M.S. Thesis. Univ. of Alaska Fairbanks, 131 pp. M.S. Thesis;
Roberts, S.C. 1994. A comparative study of the rivers of Denali National Park and Preserve, Alaska, with particular reference to the Rock Creek Inventory and Monitoring Program. M.S. Thesis. Univ. of London. 35 pp. + appendices.
Recent Submissions and In Preparation
Oakley, K., E. M. Debevec, and E. A. Rexstad. 1999. Development of a long-term ecological monitoring program in Denali National Park and Preserve, Alaska, USA. Pp. 00-00 in Proceedings, North American Symposium: Toward a unified framework for inventorying and monitoring forest ecosystem resources. Guadalajara Mexico. November 1998. Abstract. Manuscript.
Rexstad, E.A. and E. M. Debevec. Modeling temporal patterns in microtine abundance in Denali National Park and Preserve. Poster to be presented at the Ecological Society of America Annual Meeting, 8-12 August 1999, Spokane, WA. Submitted Abstract.
Presentations and Posters
1998
Oakley, K., E. M. Debevec, and E. Rexstad. 1998. Development of a long-term ecological monitoring program in Denali National Park and Preserve, Alaska. Presented at North American Symposium: Toward a unified framework for inventorying and monitoring forest ecosystem resources. Guadalajara Mexico. November 1998.
Milner, A.M. and S.J. Roberts. Classification of rivers in Denali National Park, Alaska for long-term ecosystem monitoring. North American Benthological Society Meeting, Halifax, Nova Scotia, June 1998.
Roberts, S.C. and A.M. Milner. Characterisation of streams in Denali National Park, Alaska, using macroinvertebrate communities. British Ecological Society Meeting.
Rexstad, E. Monitoring of small mammal populations in Denali National Park and Preserve. Presented at Annual Cooperators meeting, Alaska Cooperative Fish & Wildlife Research Unit, Fairbanks AK.
1997
Rexstad, E. Biological monitoring of small mammal populations in Denali National Park and Preserve, Alaska. 4th annual meeting of The Wildlife Society, Snowmass CO.
1996
Rexstad, E. Biomonitoring and small mammal population dynamics, Denali National Park. Life Sciences Seminar, Institute of Arctic Biology, Fairbanks AK.
1995
Rexstad, E. Techniques for monitoring and correlates of interannual variation of Clethrionomys rutilus in Denali National Park, Alaska. Instituto de Ecología y Evolución, Universidad Austral de Chile.
1994
Furtsch, P. R., and E. Rexstad. Survival, abundance, and reproduction in small mammal populations in Denali National Park and Preserve AK. Arctic Division AAAS, Vladivostok Russia.
Rexstad, E. Detecting differences in wildlife populations across time and space. North American Natural Resource and Wildlife Conference, Anchorage AK.
1993
Rexstad, E., and J. A. Cook. Long-term monitoring of small mammal populations in a taiga ecosystem. 78th Ecological Society of America meeting, Madison WI, August 1993.
| Having problems viewing the pdf files? Download Adobe Acrobat Reader 3.0 Now!  |