Quantitative representativeness and constituency of the long-term agroecosystem research network and analysis of complementarity with existing ecological networks

Authors: KUMAR, JITENDRA - Oak Ridge National Laboratory; Coffin, Alisa; Baffaut, Claire; PONCE-CAMPOS, GUILLERMO - University Of Arizona; Witthaus, Lindsey; HARGROVE, WILLIAM - Us Forest Service (FS)

Submitted to: Environmental Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/8/2023
Publication Date: 6/16/2023
Citation: Kumar, J., Coffin, A.W., Baffaut, C., Ponce-Campos, G., Witthaus, L., Hargrove, W. 2023. Quantitative representativeness and constituency of the long-term agroecosystem research network and analysis of complementarity with existing ecological networks. Environmental Management. 72:705-726. https://doi.org/10.1007/s00267-023-01834-9.
DOI: https://doi.org/10.1007/s00267-023-01834-9

Interpretive Summary: The USDA Long-Term Agroecosystem Research (LTAR) Network, now a decade in existence, coordinates agricultural research at a continental scale across multiple research sites. However, it is unclear how well the Network and its individual sites represent the totality of agricultural working lands in the continental USA. To better understand this, we performed a quantitative analysis of the 18 sites within the LTAR Network, based on 15 environmental characteristics, to produce maps of representativeness and constituency across the country. Representativeness shows how well the combination of environmental conditions at each location was represented by the LTAR sites, while constituency shows which LTAR site is the closest match for each location, even though the LTAR site environment may be very different. LTAR sites all include an experimental design element in their research programs, and these consist of multiple experimental areas of varying sizes. The central point of each experimental area at each LTAR site were used in the analysis, regardless of area. Since the objective of LTAR is specifically to represent agricultural working lands of three types – croplands, grazinglands, and integrated crop and grazing uses – we used maps of these areas to "mask" and remove non-agricultural areas. Our analyses were restricted to these working lands. Our results showed that LTAR representativeness was good across most of country. Representativeness for croplands was higher than for grazinglands, probably because croplands have more specific environmental criteria. Limited representativeness in the Pacific Northwest, in western coastal oak-savannah, and in the Central Valley indicate potential locations where future additional LTAR sites which could boost LTAR representativeness of working lands. Southwest mesquite shrub-scrub, coastal Atlantic wetlands, and Nebraska sandhills were more localized areas with poorer LTAR representativeness. Young, growing networks with few sites obtain the largest increases in representativeness by adding sites whose environmental conditions have large constituency areas, which we refer to as “generalist” sites. Conversely, more mature networks like LTAR gain the most incremental representativeness by adding “specialist” sites whose conditions are targeted at areas that are currently under-represented but may be smaller in size. The addition of “specialist” sites to larger networks must be guided by quantitative analyses such as this in order to be most effective. We explored the potential of boosting the representativeness of the LTAR Network by “borrowing” sites from other national networks, including the Long-Term Ecological Research (LTER) Network and the National Ecological Observatory Network (NEON). Borrowing an existing site from another network may be more efficient and effective than establishing an entirely new LTAR site. For the sites in these other national networks located in agricultural working lands, we mapped the areas for which LTER or NEON sites had higher representativeness than LTAR. Our results indicated that LTAR network representativeness could benefit from such borrowing of several NEON and LTER sites. Because such targeted cross-network collaboration could have mutual benefits for both networks, national networks might be motivated to adopt certain measurement protocols from sibling networks at selected sites in a reciprocating fashion. While this analysis exhaustively considered principal environmental characteristics related to production on working lands, we did not differentiate among LTAR sites with respect to the primary agronomic systems that they research. Nor did we consider the socio-economic context and dominant agricultural practices of the LTAR site areas. These are important aspects of socio-agroecosystems that, if considered in future analys

Technical Abstract: The USDA Long-Term Agroecosystem Research (LTAR) Network coordinates agricultural research across 18 research sites in the conterminous United States (CONUS). However, it is unclear how well these sites represent the totality of agricultural working lands within the CONUS. Therefore, we performed a quantitative analysis of the 18 sites, based on 15 climatic and edaphic characteristics, to produce maps of representativeness and constituency across the CONUS. Representativeness shows how well the combination of environmental drivers at each CONUS location was represented by the LTAR sites’ environments, while constituency shows which LTAR site was the closest match for each location. LTAR representativeness was good across most of CONUS. Representativeness for croplands was higher than for grazinglands, probably because croplands have more specific environmental criteria. Limited LTAR representativeness in the Pacific Northwest, in western coastal oak-savannah, and in the Central Valley indicate potential locations for additional LTAR sites. Southwest mesquite shrub-scrub, coastal Atlantic wetlands, and Nebraska sandhills were more localized areas with poor LTAR representativeness. We further explored the potential of “borrowing” sites from the Long-Term Ecological Research (LTER) Network and the National Ecological Observatory Network (NEON) for boosting the representativeness of the LTAR Network. Our results indicated that LTAR network representativeness would benefit from borrowing several NEON sites and the Jornada LTER site. While this analysis exhaustively considered principal environmental characteristics related to production on working lands, we did not consider the focal agronomic systems under study, or their socio-economic context. Further research integrating socio-agroecosystem context into representativeness would improve these results.