Matching Variety to Field Boosts Yield Potential

EDMONTON - Jan 28/04 - SNS -- Growers will get better results by matching the variety to their field than just their region, say Alberta Agriculture researchers.

In the past, choosing a variety for a specific location in Western Canada has been an immense challenge because of unpredictable variety performance under prairie climates. Every year, provincial departments of agriculture along with Agriculture and Agri-Food Canada, universities, producer organizations and private industry cooperate to establish regional variety trials at a number of different locations covering the major soil-climate regions in the prairies. The current practice of variety recommendation is based on the yield data from these trials that are summarized according to geographic regions.

Traditionally, varieties with the highest regional averages are recommended to local producers. However, this practice not only has little regard for the fact that not all fields in a given region have the same level of production capacity, but also ignores the reality that a variety may perform similarly even when it is grown far apart in different regions.

"Geography is not a reliable criterion for selecting crop varieties," says to Dr. Rong-Cai Yang, bio-statistician and research scientist with Alberta Agriculture, Food and Rural Development, Edmonton. "There is a need to identify sites that are similar in their yielding ability, but not necessarily contiguous in their geography, or 'isoyield' sites."

Yang has developed statistical analyses and computer software that allow identification of isoyield sites. He is using the models in a collaborative research with Dr. Stan Blade and other researchers of Alberta Agriculture's Crop Diversification Division, for the analysis of Alberta field pea variety trials that have been ongoing since 1987. The study will help farmers find 'right' varieties that are best suited to their fields.

"Our approach," says Dr. Yang, "is to stratify all test sites into isoyield groups. Sites within each isoyield group not only have similar production levels, but also differ very little in variety ranks across the sites." When different sites are grouped in a hierarchical fashion, based on levels and consistency of yield performance, a tree-like diagram called dendrogram is generated.

The second step of the new analysis is to conduct a series of statistical tests for determining a dendrogram 'cutoff' point that allows an impartial identification of different isoyield groups. These isoyield groups also form a basis for appropriately scaling down the level of the provincial field pea variety testing, a needed step to ensure its sustainable operation in the future in face of shrinking resources and a growing demand for improving the quality of variety testing.

The current pea variety selection in Alberta is based on average yields of the four regions: southern Alberta, east-central Alberta, west-central Alberta and Peace River region. The southern Alberta region is further divided into irrigated and dryland areas. The Peace River region includes some neighbouring sites in British Columbia.

Yang and his team have analyzed a total of 157 field pea variety trials conducted from 1997 to 2001 to identify isoyield groups. A considerable amount of site-to-site variation exists within each region. For example, in region 2, yields range from 757 kg/ha in Paradise Valley to 3666 kg/ha in Vegreville. Similar patterns of site variation are observed in the other four years. There is also little relationship between the geographic regions and isoyield groups. A similar number of isoyield groups are identified for 1998-2000.

"With the statistical models and software developed for the analysis of field pea variety trials now in place, we are ready to extend this work to include three major crops, barley, canola and field pea in Alberta and other Prairie provinces," says Yang. "The long-term goal of the research project is to tie production data from crop variety trials with climate and soil variation to generate isoyield maps for identifying the best match-up of varieties with their 'favored' environments and climates."

This ongoing research has been supported by Alberta Agriculture's Industry Development Sector New Initiative Fund and the Crop Diversification Division.

Yang provides ministry-wide bio-statistical services to help Alberta Agriculture researchers ensure the validity and excellence of their research. As a research scientist, he continues to conduct research to help resolve the issues of importance to producers and the industry.

The subscriber version of the article is available by Clicking here