Skip to main content
ARS Home » Plains Area » Temple, Texas » Grassland Soil and Water Research Laboratory » Research » Publications at this Location » Publication #334629
Title: Meta-analysis constrained by data: Recommendations to improve relevance of nutrient management research

Author
item EAGLE, ALISON - Duke University
item CHRISTIANSON, LAURA - University Of Illinois
item COOK, RACHEL - North Carolina State University
item Harmel, Daren
item MIGUEZ, FERNANDO - Iowa State University
item QIAN, SONG - University Of Toledo
item RUIZ-DIAZ, DORIVAR - Kansas State University

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/18/2017
Publication Date: N/A
Citation: N/A

Interpretive Summary: Five research teams received funding through the North American 4R Research Fund to conduct an analysis of the air and water quality impacts of on-farm nutrient management practices. In compiling or expanding databases for these analyses on environmental and crop production effects, researchers identified key data and reporting gaps in individual research publications from a large number of field trials in cultivated and grassland agriculture. Since the gaps created similar challenges across all five projects, the 4R Research Fund technical leadership tasked the researchers with recommending improvements in field research design, data collection, and data reporting to enhance future agro-environmental review-type analyses or “meta-analyses”. Thus the present manuscript was designed to summarize the meta-analyses as well as critical data and reporting gaps and to make recommendations for researchers and journals to aid in future review and meta-analysis efforts. The five meta-analyses projects evaluated nitrogen (N) and phosphorus (P) loss pathways and crop yield impacts, focusing on corn-based systems in the Midwest U.S. One consistent challenge was the difficulty in applying standard mathematical and statistical procedures due to system complexity and lack of relevant data. In addition, many studies did not provide tabular data (often only graphical data or summary data across treatments), measurement uncertainty estimates, or both nutrient losses and crop yield. Many studies also did not report crop N or P uptake, or their apparent recovery efficiencies. To alleviate such challenges and assist agricultural research in supporting subsequent meta-analyses, we make the following recommendations: 1) Utilize a common metadata protocol for consistent units and terminology, 2) clearly define treatments and controls, 3) provide a complete set of full-factorial measured data for each year, 4) collect and report a minimum set of auxiliary data, and 5) establish requirements for data curation and repositories in funding and publication cycles.

Technical Abstract: Five research teams received funding through the North American 4R Research Fund to conduct meta-analyses of the air and water quality impacts of on-farm 4R nutrient management practices. In compiling or expanding databases for these analyses on environmental and crop production effects, researchers identified key data and reporting gaps in individual research publications from a large number of field trials in cultivated and grassland agriculture. Since the gaps created similar challenges across all five projects, the 4R Research Fund technical leadership tasked the researchers with recommending improvements in field research design, data collection, and data reporting to enhance future agro-environmental meta-analyses. Thus the present manuscript was designed to summarize the meta-analyses as well as critical data and reporting gaps and to make recommendations for researchers and journals to aid in future review and meta-analysis efforts. The five meta-analyses projects evaluated nitrogen (N) and phosphorus (P) loss pathways and crop yield impacts, focusing on corn-based systems in the Midwest U.S. One consistent challenge was the difficulty in applying standard meta-analytical procedures due to system complexity and lack of relevant data. In addition, many studies did not provide tabular data (often only graphical data or summary data across treatments), measurement uncertainty estimates, or both nutrient losses and crop yield. Many studies also did not report crop N or P uptake (Plant total or grain), or their apparent recovery efficiencies. To alleviate such challenges and assist agricultural research in supporting subsequent meta-analyses, we make the following recommendations: 1) Utilize a common metadata protocol for consistent units and terminology, 2) clearly define treatments and controls, 3) provide a complete set of full-factorial measured data for each year, 4) collect and report a minimum set of auxiliary data, and 5) establish requirements for data curation and repositories in funding and publication cycles.