Search and Access Archived Conference Presentations

2013 ASHS Annual Conference

14808:
Modeling Environmental Parameters Affecting Processing Pea Yield in Minnesota

Tuesday, July 23, 2013
Desert Ballroom: Salons 7-8 (Desert Springs J.W Marriott Resort )
Charles L. Rohwer, Southern Research and Outreach Center, University of Minnesota, Waseca, MN
Vincent A. Fritz, University of Minnesota, Waseca, MN
Green peas (Pisum sativum L.) are a major processing crop in southern Minnesota. The University of Minnesota Southern Research and Outreach Center (SROC), located centrally in this pea processing region, has performed pea variety trials and has records of average yield from these trials dating to 2004. From 2004 to 2012, a total of 385 tenderometer-adjusted yield averages, planting dates, and harvest dates from 2 to 3 plantings per year of commercial varieties and advanced selections from vegetable breeding companies are available. The effects of multiple weather parameters (observed at the SROC) on pea yield was modeled using these data. Six environmental parameters explained 46% of the variation in yield, irrespective of plant genetics: average wind speed (m·s-1) 17 to 21 days before harvest (AWS), time to accumulate half of the total heat units (HU) and precipitation (cm) from planting to harvest (percentage of days to harvest), average soil temperature (°C) for one week after planting, average air temperature (°C) for one week before harvest, and season-long photothermal ratio (MJ m2/°C). A linear decrease in yield as AWS increased was the most substantial relationship (R2 = 0.29), suggesting that poor pollination or damage to flowers occurred under windy conditions. A simplified model could be used by processors to partially estimate yield prior to harvest. Using only AWS, day of the year at planting (1 Jan = 1), and estimated HU to maturity for each variety (available from breeding companies or past observations), 33% of the variation in observed yield was accounted for. The simplified model supports earlier planting for selections with low HU requirements, which is standard practice in the processing industry.