Title:

Nonthermal Grass Seed Cropping Systems

Investigators:

Jeffrey Steiner, USDA-ARS Agronomist
Stephen Griffith, USDA-ARS Physiologist
George Mueller-Warrant, USDA-ARS Agronomist
Lloyd Elliott, USDA-ARS Microbiologist
Donald Churchill, USDA-ARS Agricultural Engineer
Richard Dick, OSU Soil Scientist
Elaine Ingham, OSU Plant Pathologist
Paul Jepson, OSU Entomologist
Glenn Fisher, OSU Entomologist
Andrew Moldenke, OSU Entomologist
Thomas Chastain, OSU Seed Physiologist
William Young, OSU Extension
Mark Mellbye, OSU Extension
Gale Gingrich, OSU Extension

Objective 1:

Determine the effect of non-thermal grass residue management and legume rotation on pest and weed control, fertilizer use efficiency, and soil quality.

Experiment 1: Our study is being conducted at three environmentally diverse locations in the Willamette Valley, Oregon and will be used to develop principles for sustainable agriculture that are applicable to the Pacific Northwest states. Included are a well-drained but erodible site in the Silverton Hills, a moderately drained valley bottom site in Benton County, and a poorly drained seasonal wetland in Linn County. We are interested in the effects of crop establishment method (minimum versus conventional tillage), residue amount (minimal versus maximal), and crop sequence (monoculture grass versus a three-crop rotation) on soil and water quality, soil physical characteristics, and long-term cropping system sustainability.

Crop Sequences and Rotational Effects. Tall and fine fescue seed yields decreased as age-of-stand increased. There was no interaction between age-of-stand and post-harvest residue level for either crop. However, tall fescue seed yields were greater with the low straw management treatment, while fine fescue seed yields were unaffected by residue amount. Perennial ryegrass seed yields were not related to age-of-stand, but rather the year of production. Two sequences of crop age-of-stand verified this finding. The amount of post-harvest residue had no effect on seed yield and there was no interaction between age-of-stand and residue amount.

Under standard grower nitrogen fertilizer program, a first seed-year tall fescue crop following red clover does not yield any differently than a third seed-year crop of tall fescue grown in a continuous grass culture. A similar finding was found for fine fescue in rotation with red clover. However, if the nitrogen management program is changed to account for the nitrogen fixation effect of the legume seed crop, there is a benefit to the grass seed crop in the subsequent rotational sequence. When red clover grown for seed precedes a tall fescue crop, total N inputs are reduced by at least 53%. This equates to approximately $46 per acre saving in fertilizer N cost.

Under a no-till wheat production system in 1994 and 1995, the amount of tall fescue postharvest residue did not affect the yield of the following wheat crop. There was also no interaction between the year of production and amount of tall fescue residue on wheat yield. Conventional-planted wheat yielded higher than no-till wheat following tall fescue in 1995. The amount of tall fescue post-harvest residue had no effect on wheat yield and there was no interaction between planting method and residue level. Similarly, the amount of fine fescue post-harvest residue did not affect subsequent conventional planted wheat. Red clover following fine fescue was higher yielding than when following a winter fallow condition.

Nitrogen Fertilizer Management. First-year tall fescue and perennial ryegrass seed yields increased linearly with increasing total spring-applied N rates up to 107 kg ha^-1. Second seed-year tall fescue yield was optimized at 71 kg ha^-1, regardless of the amount of post-harvest residue. Third seed-year tall fescue required 71 kg ha^-1 if residue remained on the field, but if the residue was removed, optimum seed yield was achieved at 89 kg ha^-1. When 107 kg ha^-1 of N was spring applied to a third seed-year tall fescue crop with full-straw residue, seed yield was reduced by 25% compared to the optimum N rate of 53 kg ha^-1. For perennial ryegrass, optimum seed yield was obtained with 107 kg ha^-1 of spring-applied N, regardless of straw residue level.

First-year fine fescue seed yield was optimal at 53 kg ha^-1 of spring-applied N. Second-year fine fescue seed yield increased with increasing N rates up to the maximum applied of 107 kg ha^-1 N. There may not be a seed yield response to N in the second seed-year crops under full straw load with spring-applied N rates greater than 107 kg ha^-1. Optimum third seed-year fine fescue seed yield was achieved with 76 kg ha^-1 of spring-applied N if the straw was removed. With full straw load remaining, optimal seed yield was achieved with 89 kg ha^-1 of spring-applied N.

Under the poorly-drained soil conditions at the perennial ryegrass site and the moderately well-drained conditions of the tall fescue site, ammonium-N was the predominate N-form available to the crops. During the second seed-year (1994), ammonium-N levels from late-winter through late-spring averaged 10 ppm in the 0 to 10 cm soil profile, 8 ppm from 10 to 20 cm, and 4 ppm from 20 to 60 cm. For the same period, nitrate levels averaged 4 ppm from 0 to 10 cm, 3 ppm from 10 to 20 cm, and 3 ppm from 20 to 60 cm. In 1995, ammonium-N averaged 1.5 ppm for the three depths down to 60 cm, while nitrate-N averaged 0.3 ppm.

Soil pH was lower in ammonium-dominated soil systems compared to nitrate-dominated systems. The tall fescue and perennial ryegrass sites showed decreased soil pH when soil conditions maintained a high ammonium to nitrate ratio. Under cool temperatures and poorly drained soil conditions in the late-winter and spring months, high soil ammonium levels should be expected to lower soil pH when the plants are actively taking up N. This effect can occur whether or not ammonium-N fertilizer is applied. With the standard fertilizer management practices used, no differences in pH have been measured between the cropping sequence or post-harvest residue amount treatments.

For all three crops, fall-applied N did not increase crop phytomass but spring-applied N. Similarly, seed yields were responsive only to spring-applied N. A savings in fertilizer cost and application costs may be realized if it can be proven that fall-applied N fertilizer can be eliminated from these grass seed production systems. This finding is supported by other grass seed production N studies from western Oregon and by some seed grower practices. Further fertilizer N savings can result with the use of a legume seed crop in the cropping system rotation.

Volunteer Seedling Dynamics. Volunteer crop seedlings are the most common weed in perennial grasses grown for seed. The impact of volunteers depends on the relative competitiveness of the seedlings versus the established parent plants, and the potential for volunteer seedlings to achieve reproductive maturity after germinating in the fall and winter. The impact of volunteer seedlings is being investigated in all three seed crops. The experimental conditions include: low and high post-harvest residue levels, and three control methods: (1) an untreated control; (2) glyphosate sprayed in a directed band between the established crop rows; and (3) a standard grower herbicide program appropriate for each cropping system.

Volunteer perennial ryegrass seedlings that became established between the planted rows after the first-year seed crop produced seeds in the untreated control in the three years investigated to date. The volunteers contributed 10.7, 11.2, and 25.1 % of the total seed yield in 1993 (third seed year), 1994 (second seed-year), and 1995 (third seed-year). The efficacy of the standard grower herbicide program to control volunteers varied considerably between years, with volunteer seedlings contributing 3.4, 6.9, and 17.0% of the total seed yield in 1993, 1994, and 1995, respectively. In both 1993 and 1994, volunteer contributions to total seed yield in the untreated checks was two times greater in the full- than the low-residue treatrnent. However, only in 1993 with high post-harvest residue and the standard grower herbicide program did volunteer plans contributions affect total seed yield.

The directed non-selective spray eliminated volunteer perennial ryegrass and resulted in the highest seed yield of all treatments. Averaged over the three years. the standard herbicide treatment program reduced the yield of parent plants by 17%, compared with the directed row spray treatment. Competition from volunteer seedlings in the untreated check reduced total seed yield by 21 %. When the seeds produced by the volunteer plants were included in these comparisons, the reductions in total yield compared to row spraying were 10 and 8% for the standard herbicide treatment and the untreated check, respectively. There were no differences in yield between low and full-residue treatments.

Volunteer fine fescue seedlings did not become established due to low seedling vigor and the competitiveness of the established crop. Due to the creeping nature of fine fescue, there was no between-row space for volunteer seedlings to become established after the first seed-year. The standard herbicide treatment resulted in lower yields than the row spray treatment in 1995. This indicates that crop injury from the standard grower herbicide program can reduce yield. Seed yield from the untreated check in 1995 was intermediate between the directed row spray and standard grower herbicide program treatments. The three volunteer seedling treatments did not differ for the amount of general weed control efficacy in the first seed-year crop (1994). However, annual grassy weeds such as rattail fescue became a problem in the untreated check in the second seed-year crop (1995).

Volunteer tall fescue seedlings produced no seeds in 1994, despite being well-established in the untreated check. The seedlings were effectively controlled by the standard grower herbicide program and directed row spray treatments. The post-harvest residue management and volunteer control method treatments had no effect on seed yield in 1994. Seeds produced by volunteer plants contributed 10% of the total yield in the low post-harvest residue-untreated control treatments, only 1% in the high post-harvest residue-untreated control treatments, and none in the directed row spray and standard grower practice herbicide treatments. The number of volunteer seedlings were lower in the low-residue treatment, thus reducing competition with one-year-old plants and reducing recruitment of additional new seedlings following the second seed-year crop. High post-harvest residue amounts covering the one-year-old volunteer plants may have slowed their growth compared to the low residue treatment.

Slugs Populations. Slugs have been found at all three experimental sites. Slugs were inactive in the fall of 1994 until significant precipitation built up soil moisture. A single application of bait in the fall adequately reduced slug populations so that no crop damage was observed through the midwinter 1994/95 period. Slug populations began to increase again in late-winter and early-spring as temperatures increased and eggs hatched. Only one application of bait was required to control the slugs in the winter-spring of 1994, but at least two applications were needed in winter-spring 1995 because of warm and moist field conditions.

Slug populations were greater in juvenile tall fescue that was planted after red clover than in mature tall fescue stands. High slug populations were also found in plots that followed wheat that was grown the previous summer. Populations in the fall 1995 increased to very high numbers (50 to 80 slugs per bait station) due to the warm and moist conditions. Slug populations peaked by the end of December in juvenile stands of tall fescue and fine fescue, and in juvenile and mature stands of red clover. One application of bait reduced the numbers of slugs per bait station to 4 to 6. Slug numbers dropped naturally at all three locations by mid-January and remained low through March due to the cold temperatures, snow cover, and flooding. These recent weather conditions may be more indicative of future trends if the drought period has ended.

In another research area located at Hyslop Farm, very high slug populations (350 to 480 slugs/bait station) were measured in the fall 1995 following a two seed-year red clover crop. Fall-applied bait was not successful, and fall-planted perennial grasses and wheat were defoliated by the slugs. Interestingly, a late-December application of bait between rain storms resulted in nearly complete control, lowering slug populations to 15 to 25 slugs per bait station. Generally, applications during this period are reported to not be effective. Following this time, slug numbers dropped naturally at this and the three other sites by mid-January and remained low through March due to the cold temperatures, snow cover, and flooding. More observations are needed to determine the effects of crop and rotation sequences on slug populations. The amount of post-harvest straw residue appeared to have no adverse affect on slug counts made in tall fescue, perennial ryegrass, and fine fescue plots of this study.

Planned Activities for 1997:

A primary new direction for our project is to identify biological indicators that can be used to predict soil health and to decide which best management options will optimize long-term economic return, minimize agricultural chemical and energy use, and minimize negative environmental impact. To achieve this, the cooperative research team has been expanded to include Drs. Paul Jepson and Andrew Moldenke, OSU, Department of Entomology who will investigate above and below soil insect population dynamics and Dr. Elaine Ingham, OSU, Department of Botany and Plant Pathology who will look at soil microbe dynamics. To address the economics of the cropping system effects, Mr. Hal Gordon, Agricultural Economist, USDA Natural Resource Conservation Service, Portland has joined the team. Additionally, Dr. Steiner has redirected a portion of his research appointment to study weed biology in the context of the cropping systems project.