Predictable crop yield patterns within farm fields are important to farmers interested in varying management inputs within fields. Now a team of Alberta Agriculture researchers is finding that in-field yield patterns are strongly affected by site-specific landscape characteristics as well as weather conditions.
The practice of varying management inputs is often called site-specific management or precision farming. It aims to increase the economic efficiency and decrease the environmental impacts of farming. For example, if areas in a field yield poorly in most years due to weed infestations, the farmer can target specific weed control measures to those areas, rather than applying them to the whole field.
Research Agrologist Sheilah Nolan of Alberta Agriculture, Food and Rural Development is part of the research team. She says, "Yield differences by landscape are reflecting underlying differences in soil properties and water flow. So landscape differences could give us a basis for making within-field management decisions." This project is a component of a joint study with Agriculture and Agri-Food Canada’s Matching Investment Initiative Program, with funding from Norwest Labs, Agrium and Westco.
Since 1996, the researchers have measured yield differences within three fields. Each field represents a common type of Alberta landscape. One is near Hussar on strongly rolling land with a clay loam soil. The second is on a sandy loam soil in hummocky terrain near Stettler. And the third, located near Gibbons, has a sandy soil on a ridge and a clay loam in the low areas.
"At each site we located three transects from hilltop to valley bottom, for a total of about 90 points per field," explains Nolan. "Each point was located using a differential global positioning system, DGPS, so that its exact location could be found from one year to the next." From 1996 to 1998, they measured yield at each point using a plot combine. Nearby weather stations measured rainfall and air temperature.
"To test whether yield differed by landscape position, we grouped the yields according to whether they were measured on an upper (U), mid (M) or lower (L) landscape positions," she says. "We noticed that for the years with close to normal weather conditions (1996 and 1998), the yield patterns by landscape position were different between sites, but were similar within sites. However, in the dry year (1997), yield patterns were surprisingly similar for all landscape positions, at all sites."
For both "normal" years at Hussar, the yield was greatest at L and least at U, with as much as 20 bushels/acre yield difference between the two (Figure 1a). For both years at Stettler, the greatest yields occurred at M, with no statistical difference between yields from L or U (Figure 1b). These positions yielded up to 16 bushels/acre less than M. At Gibbons in 1998, the pattern was statistically different again, with U yielding the most and L the least (Figure 1c).
![](/web/20061208032613im_/http://www1.agric.gov.ab.ca/$department/newslett.nsf/37ac0bef4de4868787256bab00709a80/1634acbb96a6612d87256c4d005db837/Information/0.7C46!OpenElement&FieldElemFormat=jpg)
Figure 1. Yearly variation in crop yield at Upper (U), Mid (M) and Lower (L) landscape positions at three sites in Alberta.
The relative consistency of yield pattern by landscape position within sites during the years of normal weather conditions suggests that yield is influenced by landscape position, reflecting differences in soil properties and water movement. However, normal weather conditions only occurred for two out of three years at the Hussar and Stettler sites, and one out of two years at the Gibbons site.
"The results from the year in which drier weather conditions occurred were very surprising," Nolan notes. In July 1997, only 40% of normal rainfall fell at Hussar and only 34% fell at Stettler. For 1997, there were no statistical differences between yield at any of the landscape positions, at any of the three sites (Figure 1 a,b,c). "Although we expected to see lower yields in the U positions, instead the drier year seemed to level out yield differences, at all three sites," she says.
"It’s clear that variable management strategies require an understanding of the effects of non-normal weather conditions and that several years of field observations are needed to characterize these conditions," says Nolan. The variation in yield pattern between sites and between years emphasizes the site-specific nature of precision farming and the need for each farmer to understand the variations within each field in local climatic conditions.
For more information contact Sheilah Nolan |