Skip to Content

Trends in Long-Term Ecological Data: a collaborative synthesis project Introduction and update

Printer-friendly versionPrinter-friendly version
Issue: 
Spring 2006

- Christine Laney & Debra Peters

Introduction:

A large suite of National Science Foundation (NSF) and United States Department of Agriculture (USDA) supported long-term research sites, including the 26 LTER sites, represents a wide range of ecosystem types, from forests to grasslands and shrublands, freshwater lakes and streams, near coastal marine and estuaries as well as urban areas and systems in the arctic and Antarctica. A variety of different kinds of data have been collected from these sites through time, ranging from primarily climatic and demographic data since the 1800s to more recent quantitative assessments of plant, animal, and microbial populations and communities, hydrological and biogeochemical cycles, biodiversity, and disturbance regimes. In addition, pollen records and tree-ring data can be used to push data availability back even further. As the LTER enters its "Decade of Synthesis", the United States Forest Service (USFS) enters its "New Century of Service", the USDA Agricultural Research Service (ARS) enters the time when "The Future Grows Here", and new initiatives, such as the National Ecological Observatory Network (NEON) become operative, there is a critical need for a collection of highly accessible, up-to-date, and easy to use data sets that span the ecosystems and history of the U.S. In response to this need, the Trends project was developed as a collaborative effort between the LTER network, USDA ARS, USFS, and the National Center for Ecological Synthesis and Analysis (NCEAS).

Goals:

To create a platform for synthesis by making long-term data available, and to illustrate the utility of this platform in addressing important within-site and network-level scientific questions.

Products:

  1. A book to be published by Oxford University Press that focuses on trends in long-term data within and among sites, and examples that illustrate the value of long-term data in addressing important questions for a number of sites
  2. A web page containing derived long-term data and metadata that are easily accessible for synthetic analyses by a variety of users. In addition, the Trends project includes a NIS Data Module that was approved by the LTER Coordinating Committee. The data module, includes two main goals recommended by the IMEXEC:
    1. To develop a dynamic webpage that generates and displays plots from derived datasets
    2. To determine the feasibility of using EML in the development of NIS modules for solving general synthesis problems. The LTER Network Office (LNO) will assist in meeting these data module goals.

Aims:

  1.  
    1. Short-term (Sept. 2006): The book should be 80-90% complete, and a web page should be available to interested users in time for the All Scientists Meeting. This initial web page will contain static derived datasets, and simple search/query, download, and graphing capabilities.
    2. Long-term (2007- ): We will work with LNO to develop a Trends web page containing dynamic datasets that can be harvested from each site periodically, converted to derived form, searched, and available for download, graphing, and statistical analyses. We would also like to link with the education community through web sites such as Teaching Issues and Experiments in Ecology (TIEE, http://tiee.ecoed.net/).

Types of data being collected, and current status

Trends data and graphs will be organized into four themes that correspond to the LTER Planning Process: climate and physical variability, including disturbances; human population and economy; biogeochemistry; and biotic structure. Within each theme, the Trends Editorial Committee identified key variables useful for cross-site and network-level analyses. These variables will be plotted on monthly or annual time scales. In addition, graphs from each site will be used to illustrate the importance of long-term data in addressing ecological problems. We expect that most of these figures will already have been published. Initial cross-site synthetic analyses related to global climate change and human impacts on the environment will be investigated as part of the book.

Data sources:

  • The National Climate Data Center's website, http://www.ncdc.noaa.gov/oa/climate/onlineprod/drought/xmgrg3.html, has been an invaluable source of Palmer drought Palmer drought severity index data for all of the continental sites. Similar types of data are needed for sites in Alaska and outside of the U.S..
  • ClimDB/HydroDB: We have extensively utilized this website to locate and download derived (monthly or annual) precipitation, air temperature, water temperature, solar radiation, wind speed, streamflow, and other available data for multiple LTER and USFS sites. The ability to obtain derived data from LTER and USFS sites in the same format, with the same attribute labels, has been invaluable. This allows us to write modular plotting scripts (we currently are using R) that can be used to quickly and efficiently plot identical graphs between sites.
  • KNB and LTER Metadata catalog servers: Some data were downloaded via these servers. However, we found it difficult to search for datasets that contained 10 or more years of data using this method. We found that we also had to look at each individual site's website and/or ask the PIs or IMs directly to help us figure out which of their datasets fit the time requirements, were pertinent to the project, and contained unrestricted data.
  • Site websites: LTER websites provide a great deal of information, and more is available with each passing month. EML provides a great standard format for reading and interpreting the metadata, and EML that includes attribute-level information is the most helpful. Non-EML data is often very useful as well, but as each site uses a different layout, it will require more work to utilize these metadata files.
  • Information managers and principal investigators have often directly submitted data that was either not available on a website or in a metacat server, or was not up-to-date online. They have also submitted graphs, previously published figures, photos, text descriptions of their sites, and other useful information. The data has been submitted in various formats, ranging from simple fixed-width format to multi-page excel spreadsheets or even database files.

Current Status:

  • Participants:

    All 26 LTER sites, 9 USDA ARS sites, 14 USFS sites, and a site administered by the University of Arizona are currently participating in the Trends project. See the figure below for coverage.
  • Climate, Biogeochemistry, and Biotic Data:

    We have well over 300 datasets currently in hand, and metadata for about half of these datasets available in EML. We are currently generating some derived datasets and graphs, using R, an open source statistical program. These scripts will be stored for informational purposes, and may also be used in the future to help generate new datasets from newly harvested raw data. We are also producing EML files to document the derived data and graphs were generated. Recently, we reviewed all of the data in hand or available online to see which of the data for all of the variables are represented by each site. We have asked each site to review these data and make changes or additions to their submissions.
  • Population and Economy:

    Some of the long term data included in the Trends Project will focus on Human Population and Economy of each site. Using lists of counties representing the region associated with each LTER site, long term changes in population, economic structure (employment by type), and land use will be quantified. Counties were the chosen unit of analysis because they are the only long term geography that is available for these data. Data from 1790 to 2000 will be organized from the U.S. Census, with Chris Boone (ASU: BES & CAP) working with the data from 1790 to 1970 and Ted Gragson (UGA: CWT) using the data from 1970 to 2000. Trends in land use provided by each site will also be included. A prototype for this section was developed by Boone and Morgan Grove (USFS: BES) using data from BES.
  • Database:

    We created a database to hold basic information about each dataset, along with the locations of each dataset's data files, metadata, graphs, etc. This database will eventually be used by the LNO-developed module to allow users to access the data and metadata.

Future events:

  • Editorial committee meeting: The editorial committee will work on final formatting and organization of the book and initial web page in July.
  • All Scientists Meeting: Christine Laney and Ken Ramsey will host a working group session in the afternoon (1300 - 1600) of the ancillary day, Sept. 20. They will focus on IM issues associated with synthesis projects like Trends. Deb Peters and Christine Laney are hosting an evening session on Friday, Sept 22, to present Trends to the broader LTER community. At this point, additional data and graphs can still be added to the book and web page from the community

Trends Participants and Example Palmer Drought Severity Index graphs, 1895-2008