Data Management Approach to Multidisciplinary Agricultural Research and Syntheses Daryl E

doi:10.2489/jswc.69.6.180A

FEATURE Data management approach to multidisciplinary agricultural research and syntheses Daryl E. Herzmann, Lori J. Abendroth, and Landon D Bunderson

gricultural scientists are increas- corn-based cropping systems while work- share and analyze increasingly large and ingly asked to address challenges ing to decrease the environmental footprint diverse datasets. There is no one tech- A related to natural resource stew- under climate change. The Sustainable Corn nological solution that facilitates big ardship, agricultural productivity, and team is a partnership across 11 institutions science, and large collaboration projects environmental protection while simul- and 9 states, and includes 140 active mem- create a new set of problems traditional taneously being mindful of the impact bers from 19 disciplines working within scientists have not previously experi- and risks associated with climate change biophysical and social research, extension, enced. A key problem teams encounter (ASA CSSA SSSA 2011; Hatfield et al. and education (see table 1 in Eigenbrode is the deluge of diverse, unfamiliar datas- 2011; OSTP CAST 2012). Measuring and et al. 2014 for more details). A major ets with insufficient project preparedness predicting the effects of climate on agri- component of the proposed work was to to adequately handle them (Haendel et cultural systems adds a layer of complexity collect data from a field research network al. 2012; Michener and Jones 2012). The that is challenging using traditional data comprising existing and newly established Sustainable Corn team can be categorized management methods; these methods also experimental sites across the US Corn Belt as a big science or big interdisciplinarity Copyright © 2014 Soil and Water Conservation Society. All rights reserved. limit the potential for full data discovery while implementing standardized method- team due to the number of research sites, Journal of Soil and Water Conservation and innovation (Overpeck et al. 2011; ologies that would allow for synthesis across complexity of data, and users dependent Wolkovich et al. 2012). To properly address differing soil types, environments, and man- on the database (Eigenbrode et al. 2014). challenges, access to multidisciplinary agement operations. Our approach to data management and data spanning environments, timescales, Increasingly, proposals are encouraged project collaboration serves as a case study treatments, and management is necessary (or required) to be multi-institutional that may be a viable option for consid- (White and van Evert 2008; Reichman et and multidisciplinary with a combina- eration by other multi-institution and al. 2011; Eigenbrode et al. 2014). Disci- tion of scientists that may or may not multidisciplinary teams. plinary scientists, data scientists, and data have previously worked together. This managers need to increasingly work in a approach to funding can potentially cre- CONCEPTUAL APPROACH collaborative manner in this data-rich era. ate fragmented, disparate sets of data that No multi-institution infrastructure was in While scientists generally desire to share result in a series of publications, but do place at the time of receiving the award 69(6):180A-185A data, time constraints, limited funding, a not encourage the data to live on beyond to support project and data collaboration lackluster reward system, and reuse con- the funded team due to insufficient funds needs. Time was a precious commod- cerns are cited as barriers (Michener et or lack of professional archivists with the ity as field data collection was to start al. 2011; Tenopir et al. 2011; Marx 2012; necessary capacity or infrastructure (King immediately and the database would need Wolkovich et al. 2012). A concerted and 2011; Wolkovich et al. 2012). The type to support the next five years of effort. www.swcs.org well-executed approach is necessary to and length of funding should be consid- Also, the expectation of USDA NIFA overcome these barriers and move toward ered by teams when developing their data funded Coordinated Agricultural Projects transformative science. management plan; we believe different (CAPs) was that findings related to field The Climate and Corn-based Cropping approaches should be considered given research, synthesis, and modeling would Systems Coordinated Agricultural Project team goals, allocated dollars, existing infra- be produced during the funded period (CSCAP), referred to as “Sustainable structure, and expected use of the data and not simply toward the end, which Corn,” was funded by USDA National during and beyond the funding period. necessitated data upload, review, and man- Institute of Food and Agriculture (NIFA) The amount of agricultural data that are agement occurring concurrently and in a and is the largest corn (Zea mays L.)-based publicly available is only a fraction of the timely fashion to the collection phase to research project funded by the agency to- research conducted, which leaves much ensure use by other team members. An date (USDA NIFA 2011; Eigenbrode et al. undiscovered, not utilized by other scien- initial set of needs were identified includ- 2014). The team is investigating complex tists (current and future), and with a high ing (1) an authenticated virtual space that carbon (C), nitrogen (N), and water cycles propensity to be lost over time (White and allowed sharing of files, editing of content, to increase the efficiency and resiliency of van Evert 2008; Vines et al. 2014). and revision control; (2) a system for data “Big science” is a loosely used term entry, review, analysis, and long term stor- Daryl E. Herzmann is systems programmer, Lori to describe new scientific approaches age; and (3) a Wiki website for interactive J. Abendroth is project manager, and Landon available thanks to modern technology, editing of internal content. Although this D Bunderson is data manager for the Climate including massively scalable comput- paper provides an overview of the team’s and Corn-based Cropping Systems Coordinated Agricultural Project at Iowa State University in ing and high speed networks connecting research data approach, similar strategies Ames, Iowa. nearly every corner of the world. This and tools were utilized for overall project technology permits researchers to instantly data to have a streamlined approach and

180A NOV/DEC 2014—VOL. 69, NO. 6 JOURNAL OF SOIL AND WATER CONSERVATION Figure 1 Corp., Redmond, Washington) spread- Workflow describing the team’s research (blue shading) and project (yellow shading) sheets, while leveraging cloud-based tools databases highlighting the one-directional workflow (red lines) and expectation of team and services provided by Google with members (users) contrasted with the multidirectional workflow (grey lines and arrows) automated syncing processes written, and expectation of data personnel. The entry portal for all databases is through the when appropriate. internal team website. The workflow has been simplified in some aspects graphically. Research and project management databases utilize similar approaches and technology It is important to utilize cloud tech- although only the research component is discussed in this paper. nologies that are similar in appearance and functionality to existing software packages used for data entry. Agricultural researchers are accustomed to the Microsoft Excel style of data entry and manipulation (Hunt et al. 2001). For example, Google’s spreadsheet application is similar to Excel, yet runs within a web browser and interfaces with data within Google cloud, allowing synchronization to a server at Iowa State University (ISU) and immediate viewing Copyright © 2014 Soil and Water Conservation Society. All rights reserved. by team members. This varies substantially Journal of Soil and Water Conservation from the more traditional workflow in which data are shared with other collaborators via email or stored on some shared network file storage space. Technology alone is not the panacea of collaboration, but requires individuals (data personnel) that can build and manage the environment. A thorough comprehension of the project, workflow, and its various components allows technological solutions to be developed that directly meet 69(6):180A-185A the needs of the collectors and users of provide consistency and synergy for team faces of this data. The aforementioned the data for the project. Data team mem- members (figure 1). work does not include identity man- bers must have extensive knowledge of With members located across nine agement of collaborators, the curation the project science, data types, protocols, states, the Sustainable Corn team had to process of reviewing data, and training and methods. They should also be familiar www.swcs.org employ methods that were easy to learn, aspects of data entry. Many of these steps with the current data provenance meth- transparent, and powerful for successful are typically accomplished by a dedicated ods, tools for entry, analysis, and modelling. virtual interactions. The establishment of information technology staff member(s) Beyond having a strong technical knowl- a virtual environment to facilitate central- and perhaps a few tech-savvy power users edge, it is crucial that these individuals are ized data collection presents a number of on the project. While certainly possible, able to translate concepts and requirements challenges, which include identity man- these steps would have taken a number of in ways that users can easily understand agement, data provenance (the origin and years to complete. The desire to achieve along with having an innate ability to work curation), disparate software platforms, and the ideal technical implementation was within the full spectrum of the data cycle. data formatting. While the Internet has balanced with the immediate need of They must also be able to work within the greatly bridged the physical distance gap project members to start collecting data “grey” areas that often impede team sci- between scientists, the recent proliferation and collaborating; every effort was made ence because of an inability of others to of web browser based technologies has to accomplish the best technical approach know what is missing or needed by oth- mitigated many of the aforementioned while meeting needs of real users (Bach ers. Multidisciplinary science includes data issues with virtual environments. et al. 2011). This was the motivation types that can be messy or unfamiliar to The building out of a custom suite of behind using cloud computing, which many and must be distilled and explained tools on a local server includes the instal- in 2011, was rapidly gaining adoption as into simpler information types. Although lation and security hardening of server companies such as Google built their web technological approaches often receive the software, designing of web data entry accessible applications. The Sustainable majority of time and discussion in papers, forms, programming of these forms to save Corn team’s approach uses traditional we believe the right personnel and skillsets data to a database, quality control routines technologies, like a relational database and can be equally important, are a key indi- of this information, and download inter- conventional Microsoft Excel (Microsoft cator of success, and should be examined

JOURNAL OF SOIL AND WATER CONSERVATION NOV/DEC 2014—VOL. 69, NO. 6 181A just as thoroughly. The Sustainable Corn different formats on demand. Google’s work documents) and Google Gadgets data team members (e.g., authors of this identity management allows for accounts (customized interfaces). Using Google paper) have backgrounds in climate sci- to be used by team members to access the Gadgets, which reside on the Google ence, agronomy, biology, ecology, field plot project’s internal team site and the data- Sites Wiki pages and interface with data research, statistical packages, programming base all within the same entry portal. stored within Google Spreadsheets, we languages, technology, and management. An expansive field research network created a custom interface for all manage- These qualifications are important when was established in all project states (Iowa, ment metadata (figure 2). A feature to the determining the appropriate schema to Illinois, Indiana, Michigan, Missouri, Google Drive is data versioning, which represent collected data, the infrastructure Minnesota, Ohio, Wisconsin), except allows oversight of what data are changing to support the data collection, the ability South Dakota, and includes 35 experi- when and by whom allowing data changes to conduct quality control and checks on mental sites including long-term and to be rolled back if necessary. This is a very data entered, and the needs of others on newly established research plots. As important feature that some data manage- the team for data. described in Kladivko et al. (2014), proto- ment systems do not support. cols were developed by team scientists and In general, team members utilize the DEVELOPMENT encompass treatments within five catego- following workflow when interfacing An initial estimate of data collection vol- ries (tillage, crop rotation, drainage water with the team’s “Climate and Cropping umes determined there would not be any management, N fertilizer management, Systems” database (see figure 1): Copyright © 2014 Soil and Water Conservation Society. All rights reserved. technical limitations to storing these data and landscape position) and measured 1. Team members enter data into a cus- Journal of Soil and Water Conservation on the cloud with a redundant, noncol- data spanning agronomic, soil, water, tomized Google Gadget (webform) or laborative copy stored on local preexisting greenhouse gas, integrated pest manage- site-specific generated Google Drive infrastructure at ISU. While the automated ment, and weather. The establishment and spreadsheet. Unstructured data sets are sensor deployments common within use of standard sampling methodologies uploaded to Google Drive for manual environmental data sets will collect up across research sites and personnel helps to processing by the data manager. to a few million observations, the total minimize variations, ensure scalability, and 2. Data are gathered, processed, and quality size of Sustainable Corn data will only allows for transparent management of data. controlled within the ISU housed data- reach a few gigabytes. While size was not The worth and value of standard methods base. Data are synchronized between a constraint, the complexity came in the in agricultural research has been recog- the ISU database and Google Drive. diversity and methods by which the data nized, and there is great potential for more 3. Team members who wish to use the were collected. The implementation of widespread utilization (Hunt et al. 2001; data may request it via Google Gadgets 69(6):180A-185A accounting for this complexity was not Tenopir et al. 2011; Olson 2013; White (webforms) that materialize data from appealing to accomplish via the genera- et al. 2013). Team protocols have helped the ISU database and/or directly from tion of data entry software on the local to ensure consistency in data collected the Google Spreadsheets. ISU data server. Instead, we determined among sites; accountability of data origi- To allow simple, straightforward, and customized entry cloud spreadsheets and nating from each site; and a consistent set transparent data entry portals that also www.swcs.org interfaces would best handle the complex- of data for systems analysis and predictive force uniformity and structure, we began ity with the option of writing software modeling across a suite of local, regional, by dictating the exact combination of against the data stored which provides a and national scale models. treatments and data variables that would level of abstraction to the data within The management of project research originate from each site over the course the project itself. Others are free to write data consists of a hybrid use of Google of the five years. Two relational spread- code to interface with the web services collaboration tools along with tradi- sheets were developed to encompass the to manipulate the data just as the official tional relational database software. For a complexity of the field research network Sustainable Corn project interfaces do. discussion of relational databases for agro- using unique site identifiers and coding Today’s web browser allows for cen- ecological research, see van Evert et al. systems. One is a master treatment matrix tralized identity management (session (1999). Through the use of Google web encompassing all treatments (55), and the cookies), data provenance through trans- services, these data repositories are auto- other is a master data matrix including all actional web services tracking access matically synced via custom software to the data variables (95). Although consistent, and changes, the standardization of web local database at ISU. This syncing limits standardized protocols were developed, applications using HTML/Javascript issues with data versioning and allows data variations existed in the specific treatments technologies, and the development of providers to manipulate and data users to or data collected due to differing capaci- community data formats and transports. download data through customized inter- ties across the 35 research sites. Codifying Web browser users are able to seamlessly faces. The manipulation and download these protocols into database schema and pass between federated websites using a interfaces are organized within an internal metadata was completed in the first year. single identity, modify shared data stored website (Google Sites) and accomplished Data catalogs documenting standards, in the cloud through no-software-installed through two mechanisms: Google Drive units, and explanations for the metadata, applications, and export this data in many (collaboratively edited spreadsheets and treatments, and measured data were devel-

182A NOV/DEC 2014—VOL. 69, NO. 6 JOURNAL OF SOIL AND WATER CONSERVATION Figure 2 International Benchmark Sites Network Screenshot of the web interface by which collaborators enter, edit, and delete for Agrotechnology Transfer (IBSNAT) field management information (metadata). This is an example of the field opera- and the Agricultural Information tions tab entry webform for a research site in the 2013 season. This application is a Management Standards (AGROVOC) Google Gadget, which resides within a Google Site, and saves the data to a Google (Hunt et al. 2001; FAO 2012; ASA CSSA Spreadsheet. The code for the interface is written in Javascript and HTML. SSSA 2013). We currently have approxi- mately 100 metadata types that can be entered per research site. This Google Gadget interface populates into and reads out from a Google Spreadsheet that can then be related to field data. Metadata can be entered, edited, or deleted by team members in real-time via the Google Gadget. Python (van Rossum 1990) scripts have been developed to interface with these spreadsheets via Google’s GDATA Application Programming Interface (API). Copyright © 2014 Soil and Water Conservation Society. All rights reserved. These scripts download the raw data from Journal of Soil and Water Conservation the Google Spreadsheets and mine them into a traditional relational database for aggregation, quality control, and backup purposes. The Google Drive API also provides metadata on spreadsheet changes that allows the download scripts to effi- ciently process only new data. Datasets generated via automated methods or continuous (such as weather data, moisture sensor readings, etc.) are uploaded directly to the Google Drive and stored in 69(6):180A-185A their native format, be it a structured text file or binary data. Processing scripts are then built to download these files when the Google Drive API advertises them as new or modified. These scripts process the raw www.swcs.org data into the local database. With the data centrally collected within the Google cloud and relational database, the third step was to provide download interfaces that collated and presented data in a user-friendly format. Again, Google oped to ensure all are clearly interpreted into these spreadsheets are the plot identi- Gadgets were chosen to provide metadata- by team members. fiers as these can be numerically sorted and driven and dynamically generated website To create customized entry spread- align directly with their Excel spreadsheets forms that present users with download sheets that aligned with users’ Excel files, allowing them to copy and paste. options to select from the diverse datasets each treatment occurring at a particular Metadata detailing site management collected. These forms call web services site was assigned a unique plot identifier by operations are entered through a custom- that package the data into a format that is the user; this plot identifier was typically a ized interface which prompts members importable into a desktop spreadsheet pro- number sequence identical to that located with a menu of options that are populated gram or statistical software package. At any on the plot stake. These plot identifiers are with drop-down lists and open text boxes point, users are free to directly interrogate then combined with the master treatment where information pertaining to land the Google Spreadsheets and manually and master data matrices to generate cus- management, treatments applied, and prac- collect the data they need as it is a com- tom entry spreadsheets that only show the tices important for data interpretation are pletely accessible and transparent system. viable treatment and data combinations entered (figure 3). Existing standards have for a particular site. The primary sorting been incorporated as applicable, such as criteria for field personnel when entering use of SI units and those stated within the

JOURNAL OF SOIL AND WATER CONSERVATION NOV/DEC 2014—VOL. 69, NO. 6 183A IMPLEMENTATION easily see where their data input may be a fairly robust and functional system run- Our strategy has been a phased approach, lacking. Also, a yearly audit is made on each ning very quickly. releasing entry interfaces and spreadsheets spreadsheet, and the data are reviewed with The team’s data management approach to the team as they are finalized (following researchers who have responsibility for and research database (Climate and testing) to allow data upload and increase uploading data. This is essential to ensure Cropping Systems database) has provided user familiarity and ownership. With a missing data are not still pending, data are structural and collaborative advantages for phased approach, it was crucial to get the in the correct units, and all required data team members and working groups. By the data entry portal live as entry and qual- have been collected. The audit also affords end of the project (2016), we believe this ity control are the most time consuming the opportunity for researchers and the approach will be documentable in terms activities; this potential bottleneck had to database team to review requirements and of increased collaboration, synthesis, and be avoided to ensure project goals were give feedback to team leadership. In addi- greater deliverables (such as publications) met. The entry portals (metadata and field tion to yearly audits, the database team is to our funding agency. Positive outcomes data) were complete in years 1 and 2 (2011 in constant contact with each researcher as of the team’s approach encompass the data and 2012) of the project, and the download data are coming in. In order for datasets to and research cycles including the following: interfaces, similar to the customized entry be used by the modeling team and future • Standardization and decoding of soil, interfaces, were completed in year 3 (2013). researchers, metadata have to be extremely water, and crop datasets for greater Tracking the inflow of data is essential detailed, thus there is a requirement for application across disciplines. Copyright © 2014 Soil and Water Conservation Society. All rights reserved. in order for the database team to ensure constant communication. • Expedited discovery of relevant project Journal of Soil and Water Conservation uniformity and quality of data, compliance The combined approach of using a data through integrated search pro- of research team members, and proper traditional relational database in tandem vided by the Google cloud platform. storage of all data types. The database has with cloud technologies has been very • Minimal loss of data and supporting two virtual dashboards where team mem- positive overall from a data management information due to centralized storage bers can view which data were collected, perspective as well as for users. There are and metadata assigned to data. which were marked as “did not collect,” technical nuances when working with a • Improved transparency and reproduc- and which data are not yet entered into third-party free software (Google) that do ibility of findings as data are centrally the system (figure 3). require some time and attention by data- located for all team members. Dashboards are invaluable when man- base personnel. The cloud is rapidly being • Increased speed and mobilization of the aging the inflow of data. As data upload developed, however, and the benefits of team when addressing these emerging deadlines approach, team members are leveraging this have far outweighed these issues or grand challenges. 69(6):180A-185A directed to view the dashboard and can minor irritations and allowed us to have

Figure 3

Screenshot of the data status dashboard per research site and measured variable with the number of data points denoted. The www.swcs.org example here has been simplified by eliminating research site and variable identifiers. This web interface dynamically gener- ates mini progress bars based on the amount of data uploaded versus the amount of data expected, which is derived from entered metadata. Four status options exist per variable including: data have been entered (green), missing data (blue), did not collect (yellow), and not entered yet (red).

184A NOV/DEC 2014—VOL. 69, NO. 6 JOURNAL OF SOIL AND WATER CONSERVATION Hatfield, J.L., K.J. Boote, B.A. Kimball, L.H. Ziska, Reichman, O.J., M.B. Jones, and M.P. Schildhauer. ACKNOWLEDGEMENTS R.C. Izaurralde, D. Ort, A.M. Thomson, and 2011. Challenges and opportunities of open data The authors greatly appreciate the feedback and D. Wolfe. 2011. Climate impacts on agri- in ecology. Science 331:703-705, doi:10.1126/ willingness to explore this new approach to data culture: Implications for crop production. science.1197962. management by Sustainable Corn team mem- Agronomy Journal 103:351-370, doi:10.2134/ Tenopir, C., S. Allard, K. Douglass, A.U. Aydinoglu, bers. The team’s centralized Climate and Cropping agronj2010.0303. L. Wu, E. Read, M. Manoff, and M. Frame. 2011. Systems database has been developed for a regional Hunt, L.A., J.W. White, and G. Hoogenboom. 2001. Data sharing by scientists: Practices and per- collaborative project supported by the USDA Agronomic data: advances in documentation ceptions. PLoS ONE 6(6):1-21, doi:10.1371/ National Institute of Food and Agriculture (NIFA), and protocols for exchange and use. Agricultural journal.pone.0021101. Award No. 2011-68002-30190, Cropping Systems Systems 70(2-3):477-492, doi:10.1016/ USDA NIFA (USDA National Institute of Food and Coordinated Agricultural Project: Climate Change, S0308-521X(01)00056-7. Agriculture). 2011. NIFA announces grant to study Mitigation, and Adaptation in Corn-based Cropping King, G. 2011. Ensuring the data-rich future of the effects of climate change on agricultural and for- Systems. Project website: sustainablecorn.org. The the social sciences. Science 331:719-721, est production. Washington, DC. http://www.csrees. 11 institutions comprising the project team include doi:10.1126/science.1197872. usda.gov/newsroom/news/2011news/02181_cli- the following land grant universities and USDA Kladivko, E.J., M.J. Helmers, L.J. Abendroth, D. mate_change_cap.html). Agricultural Research Service (ARS): Iowa State Herzmann, R. Lal, M. Castellano, D.S. Mueller, van Evert, F.K., E.J.A. Spaans, S.D. Krieger, J.V. Carlis, University, Lincoln University, Michigan State J.E. Sawyer, R.P. Anex, R.W. Arritt, B. Basso, J.V. and J.M. Baker. 1999. A database for agroecologi- University, The Ohio State University, Purdue Bonta, L. Bowling, R.M. Cruse, N.R. Fausey, J. cal research data: II. A relational implementation. Copyright © 2014 Soil and Water Conservation Society. All rights reserved. University, South Dakota State University, University Frankenberger, P. Gassman, A.J. Gassmann, C.L. Agronomy Journal 91:62-71, doi:10.2134/agron Journal of Soil and Water Conservation of Illinois, University of Minnesota, University of Kling, A. Kravchenko, J.G. Lauer, F.E. Miguez, E.D. j1999.00021962009100010010x. Missouri, University of Wisconsin, and the USDA Nafziger, N. Nkongolo, M. O'Neal, L.B. Owens, van Rossum, G., et al. 1990. The python language ARS in Columbus, Ohio. P. Owens, P. Scharf, M.J. Shipitalo, J.S. Strock, reference. Python Software Foundation. http:// and M.B. Villamil. 2014. Standardized research docs.python.org/py3k/reference/index.html. REFERENCES protocols enable transdisciplinary research of Vines, T.H., A.Y.K. Albert, R.L. Andrew, F. Débarre, ASA CSSA SSSA (American Society of Agronomy, climate variation impacts in corn production D.G. Bock, M.T. Franklin, K.J. Gilbert, J-S. Moore, Crop Science Society of America, and Soil systems. Journal of Soil and Water Conservation S. Renaut, and D.J. Rennison. 2014. The availabil- Science Society of America). Climate Change 69(6):532-542, doi:10.2489/jswc.69.6.532. ity of research data declines rapidly with article Position Statement Working Group. 2011. In Marx, V. 2012. My data are your data. Nature age. Current Biology 24:94-97, doi:10.1016/j. Position statement on climate change. Working Biotechnology 30(6):509-511, doi:10.1038/ cub.2013.11.014. 69(6):180A-185A Group Report. Madison, WI: American Society nbt.2243. White, J.W., and F.K. van Evert. 2008. Publishing of Agronomy, Crop Science Society of America, Michener, W.K., S. Allard, A. Budden, R.B. Cook, K. agronomic data. Agronomy Journal 100:1396- and Soil Science Society of America. Douglass, M. Frame, S. Kelling, R. Koskela, C. 1400, doi:10.2134/agronj2008.0080F. ASA CSSA SSSA. 2013. Publications handbook Tenopir, and D.A. Vieglais. 2011. Participatory White, J.W., L.A. Hunt, K.J. Boote, J.W. Jones, J. and style manual: Chapter 7. In Units and mea- design of DataONE – Enabling cyber infrastruc- Koo, S. Kim, C.H. Porter, P.W. Wilkens, and surements. https://www.agronomy.org/files/ ture for the biological and environmental sciences G. Hoogenboom. 2013. Integrated descrip- www.swcs.org publications/style/chapter-07.pdf. 11:5-15, doi:10.1016/j.ecoinf.2011.08.007. tion of agricultural field experiments and Bach, K., D. Schäfer, N. Enke, B. Seeger, B. Michener, W.K., and M.B. Jones. 2012. Ecoinformatics: production: The ICASA Version 2.0 data stan- Gemeinholzer, and J. Bendix. 2011. A com- Supporting ecology as a data-intensive science. dards. Computers and Electronics in Agriculture parative evaluation of technical solutions for Trends in Ecology and Evolution 27(2):85-93, 96:1-12, doi:10.1016/j.compag.2013.04.003. long-term data repositories in integrative biodi- doi:10.1016/j.tree.2011.11.016. Wolkovich, E.M., J. Regetz, and M.I. O’Connor. versity research. Ecological Informatics 11:16-24, Olson, K.R. 2013. Soil organic carbon seques- 2012. Advances in global change research doi:10.1016/j.ecoinf.2011.11.008. tration, storage, retention and loss in U.S. require open science by individual research- Eigenbrode, S., L. Wright Morton, and T. Martin. croplands: Issues paper for protocol development. ers. Global Change Biology 18:2102-2110, 2014. Big interdisciplinarity to address climate Geoderma 195-196:201-206, doi:10.1016/j. doi:10.1111/j.1365-2486.2012.02693.x. change and agriculture: Lessons from three geoderma.2012.12.004. USDA Coordinated Agricultural Projects. Journal OSTP CAST (Office of Science and Technology of Soil Water Conservation 69(6):170A-175A, Policy: President’s Council of Advisors on doi:10.2489/jswc.69.6.170A. Science and Technology). 2012. Report to the FAO (Food and Agriculture Organization of the President on agricultural preparedness and the United Nations). 2012. Agricultural informa- agriculture research enterprise. Washington, DC. tion management standards: AGROVOC. http:// http://www.whitehouse.gov/sites/default/files/ aims.fao.org/standards/agrovoc. microsites/ostp/pcast_agriculture_20121207.pdf. Haendel, M.A., N.A. Vasilevsky, and J.A. Wirz. 2012. Overpeck, J.T., G.A. Meehl, S. Bony, and D.R. Dealing with data: A case study on information Easterling. 2011. Climate data challenges in the and data management literacy. PLoS Biology 21st century. Science 331:700-702, doi:10.1126/ 10(5):1-4, doi:10.1371/journal.pbio.1001339. science.1197869.

JOURNAL OF SOIL AND WATER CONSERVATION NOV/DEC 2014—VOL. 69, NO. 6 185A