Supply Chains and Barley Breeding

Improving the cost effectiveness of crop improvement research requires actionable information that better aligns the efforts of crop breeders with the changing demands of the producers and, ultimately, consumers they serve.

The Changing Demand for Malting Barley

Craft beer production volume (2001-2017)

The demand for U.S. malting barley is in the midst of a major structural change. In 1990 there were 284 breweries in the United States. Just seventeen years later, in 2017, that figure jumped to 6,372 breweries, almost all due to the growth of micro-breweries and brew pubs who now account for almost 30 percent of total U.S. beer production.

Hand in hand with the dramatic change in the structure of the brewing industry has been a change in the attributes consumers’ demand of the barley used to produce the beer they drink. Increasingly consumers are seeking barley that

  • is sourced locally (not shipped in from elsewhere in the country or elsewhere in the world)
  • is grown under environmentally friendly conditions (e.g., that minimizes nutrient run-off which diminishes water quality in lakes and streams)

In turn, these changing market realities are changing the realties faced by barley breeders who must now maintain the malting quality (and quantity) of grain while adjusting to the new biotic (pest and disease), abiotic (environmental) and crop management realties that come with growing different varieties, in different locations, under different production conditions.

The Agroinformatics of Barley Breeding

G.E.M.S is working with Kevin Smith (UMN’s barley breeder) and breeders from other states to better target their barley breeding efforts, cognizant of these new, and still evolving, market demand realties. This involves generating information products that support decisions along the entire barley innovation supply chain. Making integrated decisions along this entire supply chain requires linking crop breeding choices—involving access to and the analysis of G (genomics), E (environment) and M (management) data—with varietal production and crop processing choices based on S (socioeconomic) information.

The G.E.M.S scope of work—some of which is under way, some awaiting funding—includes

  • Processing geocoded statistical data and producing new remote-sensed data to geographically and environmentally map the changing location of barley production
  • Developing geo-referenced information on “market sheds” for barely, linking the evolving location of brewers and their barley use with the location of barley production
  • Assessing the biotic and abiotic “representativeness” of present U.S. trial sites for barley breeding with the present and prospective location of barley production
  • Making G x E x M trial-site data interoperable across multiple locations and multiple years to accelerate crop modeling and improve crop breeding progress in ways that optimally match released barley varieties to the local conditions under which they are grown
  • Spatially and temporally linking micro- and field-scale barley trial and farm production data with meso-scale climate, terrain, soil and water-body data to inform choices that affect the productivity and environmental performance on new barley breeds.

Localizing Barley Breeding

G.E.M.S is building out a suite of modularized, increasingly-more powerful analytical tools that help breeders optimize the choices that make in terms of which genetics to test, in which locations, for which season (winter versus spring), and under what crop management strategies. The overarching objective is to provide an on-going stream of actionable information to inform the optimal development and deployment of new barley varieties given the spatially and temporally variable market and environmental realities faced by farmers.

One module in the GEMSTool kit enables barley breeders to assess the environmental match of their trial sites to present and prospective barley production areas. With this information, they can assess either a) the commercial barley production areas that best align with the environmental attributes of existing trial sites, or, alternatively, b) how they might relocate their trial sites (or restructure the trials at those sites) to better represent the target areas in which their new varieties are to be (locally) deployed.


Enviro-matching Barley Trials

Workflow – Enviro- and market-matching barley trials

Enviro-matching trial sites requires a series of steps. First, breeding sites are carefully georeferenced and standardized so that the same geographic projection, datum and data type are used for all sites. Based on expert advice, a set of environmental variables deemed important for the performance of the targeted crop and/or variety is selected and checked for collinearity. Environmental variables/predictors that are found to be highly correlated are either removed from the list of predictors or a reduced dimension that represents a suite of variables is used for the enviro-matching. A set of statistical and machine learning models are then parametrized and trained on selected environmental data at the breeding sites so that a spatial projection showing the likelihood of any location being environmentally similar to the breeding sites is generated.

Market-matching Barley Trials

The enviro-matching results suggest that the northern parts of Minnesota have the closest environmental match to the present U.S. barley breeding sites, whereas the figure below reveals that most of the commercial barley production in Minnesota takes place in the northwestern part of the state. Further insights come from overlaying additional geospatial data layers. For example, in the figure below, the barley environmental match grid is spatially overlaid with year 2016 barley cropping data plus land cover data showing areas where shrublands and forests are found. Such geospatial overlays provide insights into areas that may have the potential to successfully produce barley but are presently planted to other crops or in areas outside agriculture.

Linking barley production to crop breeding