Abstract
The Columbia Basin Agricultural Research Center (CBARC) is home to the oldest experiments in the Pacific Northwest (PNW). The perennial grassland, continuous cereal, and residue management experiments were initiated 1931, and the tillage-fertility, wheat-pea rotation, and no-till wheat were initiated in 1940, 1963, and 1982. This article summarizes the results obtained in 2002 and 2003. The perennial grassland serves as a base-line for comparisons with other systems. Continuous cereal: in both conventional and no-till cropping systems, spring barley produced the highest yield followed by winter wheat and then spring wheat. In winter wheat cropping systems, there was no difference in yield between conventional and no-till systems. Spring wheat and spring barley produced higher yields under conventional tillage than under no-till cropping system. Crop residue: high yields were obtained when manure or nitrogen (N) was applied. Field burning without N application resulted in the lowest yields. Wheat-pea rotation: for wheat, highest yields were produced when plots were plowed in the fall or spring. The lowest yield was produced in the no-till system where downy brome infestation was significantly high. For peas, the highest yield was produced under the no-till system and the lowest yield was produced when plots were plowed in the fall. Tillage fertility: increasing N rates up to 120 lb/acre increased yield through greater numbers of heads/ft2. The highest yield was produced under the plowed plots, which had a greater number of heads/ft2 than the disked and swept plots. The sweep treatment produced the lowest yield, probably because of a high downy brome infestation.No-till wheat: in both no-till treatments, grain yields increased with increasing N fertilization up to 80 lbs/acre, then declined. There was little difference between the older set of plots (A) and the newer set of plots (B). Grain yields in the conventional tillage plots, with the same N fertilization rates, were consistently greater. Residue yields showed similar trends as the grain yields.
Introduction
Long-term research guides future agricultural development by identifying the effects of crop rotation, variety development, fertilizer use, aerial and surface contamination, and organic amendments on soil productivity and other beneficial soil properties. Comprehension and evaluation of many changes often requires 10-20 years to identify and quantify. Soil microflora and soil-borne plant pathogens require from 2 to 8 years in a new cropping sequence or tillage system to reach a stable equilibrium. To this end, long-term experimentation is required to understand interactions among soil, water, and plant factors for both agronomic and agricultural policy decisions. The oldest experiments in the PNW are at CBARC, Pendleton, in the intermediate rainfall zone (Table 1). Below is a brief description of these experiments and the results obtained in the 2002 and 2003 crop years. The treatments have changed over the years and the descriptions below refer to current procedures. Detailed descriptions of the protocols and how they have changed over time have been compiled into a database located on our network server. A detailed progress report on these trials is planned for later this year and will contain both detailed descriptions of each experiment and more extensive analysis of the trials over time. During 2003, data collected from 1998 through 2003 were reviewed and data summaries for all years of all experiments were prepared.
Description of Experiments
Perennial grassland
The perennial grassland site (150 ft wide by 360 ft long) contains no experimental variables, but has been maintained since 1931. The site is intended to approximate a near-virgin grassland and serves as a base-line for evaluating changes in other cropping systems. It is periodically reseeded with introduced grass selections, occasionally fertilized, and infrequently irrigated. The dominant grass species are bluebunch wheatgrass (Agropyron spicatum var. 'Secar') with lesser amounts of Idaho fescue (Festuca idahoensis var. 'Joseph'). Weeds, particularly witchgrass (Panicum capillare), common mallow (Malva neglecta), and downy brome (Bromus tectorum), are controlled as needed. This site received limited grazing from 1931 to 1985. It has not been grazed since, but vegetation is sometimes clipped during or after summer growth. Above-ground productivity has been measured since 1996. This area has recently gone through a renovation process involving repeat applications of glyphosate to kill existing species, very shallow tillage, and reseeding with a John Deere (JD) power-till drill.
Continuous cereal
The objectives of the various continuous cereal monocultures have varied over the years; however, the current objective is to determine the effects of annual mono-cropping on crop yield and soil productivity. Annual monoculture plots of winter and spring wheat and spring barley, using plow (inversion) tillage are maintained. In each plot there are fertilized and unfertilized blocks. Treatment histories for the tilled plots are shown in Table 2. A no-till (direct seeded) annual winter and spring wheat and spring barley companion plot was established in 1998 and the treatments are shown in Table 3. The plots are not replicated. The most practical, generally recommended methods and equipment available to growers are used. In 2002 and 2003, a JD 8300 double disk drill on 6.8-inch spacing was used to seed all conventional till monocultures. In 2002, a JD 1560 disk drill on 7.5-inch spacing and a Conservapak (CP) hoe drill on 12-inch spacing were used to seed no-till spring barley and no-till winter wheat plots. No-till spring wheat plots in 2002 were seeded with the JD 1560 drill. In 2003, all no-till plots were seeded with the CP drill. In both 2002 and 2003, all spring barley plots were seeded to 'Baronesse'. Spring wheat plots were seeded to 'Alpowa' in 2002 and 'Zak' in 2003. All winter wheat plots were seeded to 'Stephens' in 2002. In 2003, no-till winter wheat was seeded to 'Clearfirst' and conventional till winter wheat was seeded to 'Stephens'. In 2002 and 2003, all fertilized monocultures received the equivalent of 100 lbs/acre of 16-20-0-14 (N, P, K, S). In conventional plots this was applied as a plowdown dry product and in no-till plots this was drill applied either as a liquid or as a dry product. In conventional till monocultures the balance of the nitrogen (N) was applied as plowdown urea and in no-till monocultures the balance of the N was drill applied as urea granules or urea-ammonium nitrate solution.
Crop residue management
The Crop Residue experiment is the most comprehensive of the long-term experiments at Pendleton. The objective of the experiment is to determine the effects of N application, burning, and pea vine and manure application on soil properties and productivity in a conventional moldboard plow, winter wheat-summer fallow production system. Treatment history is shown in Table 4. The experimental design is an ordered block consisting of nine treatments (10 originally) and two replications. The experiment contains duplicate sets of treatments that are offset by 1 year so that data can be obtained annually. In 2002 and 2003, plots were seeded to 'Stephens' using a JD 8300 double disk drill on 6.8-inch spacing. Inorganic N was supplied as urea ammonium nitrate applied preplant using a shank applicator.
Tillage fertility
The objective of the Tillage Fertility experiment is to determine the effects of three tillage regimes and six N rates on soil properties and productivity in a tilled winter wheat-summer fallow production system. Treatments are shown in Table 5. The experimental design is a randomized block split-plot, with three replications. Main plots consist of three primary tillage systems (moldboard plow, offset disk, and subsurface sweep) and subplots of six fertility levels. In 2002 and 2003, plots were seeded to 'Stephens' with a JD 8300 double disk drill on 6.8-inch spacing.
Wheat/pea
The wheat/pea experiment was established in 1963. The objective of the experiment is to determine effects of four different tillage regimes on soil properties and productivity in a wheat/legume annual crop rotation. Treatments are shown in Table 6. Crop rotation is winter wheat/dry spring pea and the experimental design is a randomized block with four replications. Each replication contains eight plots (four treatments duplicated within each replication). Duplicate treatments, offset by 1 year, ensure yearly data collection for both wheat and peas. In 2002 and 2003, all tilled plots were seeded using a JD 8300 double disk drill on 6.8-inch spacing. In 2003, no-till peas were sown using a Great Plains double disk drill on 10-inch spacing and no-till wheat was sown using a Noble split packer drill on 10-inch spacing. In 2002 all no-till plots were sown using a JD 1560 disk drill on 7.5-inch spacing. In both years, 'Stephens' winter wheat and 'Universal' dry pea were sown. All fertilizer was applied as pre-plant shank-applied liquid fertilizer. Tilled winter wheat plots received 80 lb N/acre and no-till winter wheat plots received 90 lb N/acre. All pea plots received 16 lb N/acre. Both peas and wheat receive P and S along with the N application.
No-till wheat (summer fallow)
This experiment was established in 1982 and last revised in 1997. The modifications made in 1997 offered an opportunity to make comparisons between new and established direct seed systems. Since the fall of 1997, the overall experiment has consisted of three different components: (1) a 20-year-old no-till management system, with five N levels; (2) new treatments incorporating a 5-year-old no-till management system, also with five N levels; and (3) another 5-year-old addition utilizing conventional tillage, with only two N levels. The three main objectives of these components are: (1) to determine if any significant changes in soil quality occurred in the older portion of the experiment after 20 years of direct seeding (NT); (2) to evaluate the rate of change in selected soil parameters with adoption of NT; and (3) to identify problems that may occur during a transition from conventional tillage to direct seed and to mitigate those adverse changes. The experiment was designed so that half the plots are cropped and half are fallow in any given year, with the subsequent year cropping system reversed, thus allowing yield data to be taken every year. The experiment (with the exception of the tilled component) is strictly without tillage, other than during seeding and stubble flailing. Plots are usually seeded in mid-October with 'Stephens' wheat using a modified Noble no-till drill using HZ openers on 10-inch centers. Seeding rate is normally in the range of 105 to 110 lbs seed/acre and N is added as Solution-32, with P and S also banded at seeding (Table 7). Herbicides are used to control weeds in both fallow and crop no-till plots and the conventionally tilled plots are rod weeded.
Results and Discussion
Precipitation and temperature
The Pendleton station received 82 percent and 98 percent of 71- and 72-year average crop-year precipitation in 2002 and 2003, respectively (Table 8). Winter precipitation amounted to 81 percent and 106 percent of 71- and 72-year average winter precipitation in 2002 and 2003, respectively. Spring precipitation was 83 percent of the 71-year average in 2002 and 85 percent of the 72-year average in 2003. Based on growing degree days (GDD), the crop-year and winter temperatures were slightly warmer than the 71- and 72-year average in both years (Table 8). The spring was cooler than the 71-year average in 2002 but warmer than the 72-year average in 2003.
Managed perennial grassland
This perennial grassland serves as a base-line for comparisons with other systems. Usually scientists sample the area to obtain data to answer specific questions they are investigating at other sites. There is no systematic data collection from the grassland. Limited data are available since 1996 when above-ground biomass of the grasses was measured. Soil data are collected every 5 and 10 years to determine soil carbon status of the grassland. This area has recently gone through a renovation process involving repeat applications of glyphosate to kill existing species, very shallow tillage, and reseeding with a JD power-till drill. Following renovation, protocols have been established for continuing biomass measurement and soil data collection.
Continuous cereal
Plant stand
Plant counts, to determine plant stand, were started in 2003 (Table 11). In that year only about 50 percent of the target stand was achieved in conventionally tilled winter wheat plots compared to about 80 percent in the no-till winter wheat. The conventional winter wheat was "dusted in" because 2002 fall rains fell late. The first significant rain was on October 29th with below normal precipitation during November of 2002. Dusting seed in usually results in poor plant stands because seeds are either placed too deep or between loose clods where conditions are not conducive for maximum water uptake (imbibition). In contrast, direct seeded (no-till) plots did not have these problems and higher plant stands were achieved. In both conventional and no-till winter wheat plots plant stands were higher in unfertilized plots compared to fertilized plots.
Moist soil conditions in the spring of 2003 allowed more than 70 and 80 percent of the target stands to be achieved in spring wheat and spring barley, respectively (Table 11). More plants germinated in fertilized plots than in unfertilized plots of both spring wheat and barley. Plant stand was more than the target stand in no-till spring wheat and barley plots.
Grain yield and yield components
The continuous cereal cropping systems plots are not replicated and therefore combine yield cannot be statistically compared. However, it is statistically acceptable to compare the systems through t-tests conducted on four bundle samples obtained from each plot. Each bundle consisted of four drill rows, 1 m long, which were hand cut and threshed, then analyzed for straw and grain content. The bundle yields were highly correlated to combine yields (r = 0.94) and therefore inferences on bundle yields could be applied to combine yield with confidence except in a few cases. In 2002, all fertilized plots produced significantly higher bundle yields than unfertilized plots (Tables 9, 10). Combine data show that the unfertilized plots yielded 54 percent of fertilized plots. In 2003, the yield of fertilized conventional winter wheat, spring wheat, and spring barley was not significantly different from the unfertilized plots (Tables 11, 12). Under the no-till system, yields of fertilized continuous winter wheat and spring barley were significantly higher than the unfertilized plots. The yield of fertilized no-till spring wheat plots was not significantly different from the unfertilized plot.
Conventional tillage
In 2002, among the fertilized conventional tillage plots, spring barley produced significantly higher yields than winter and spring wheat (Tables 9, 10). Winter wheat produced higher yield than spring wheat although the difference was not significant. The results were similar for unfertilized plots. In 2003, results similar to 2002 were obtained when comparing fertilized plots. However, no significant differences were observed among the unfertilized winter wheat, spring wheat, and spring barley.
No-till
In 2002, the CP and the JD 1560 were both used to seed winter wheat and spring barley. In fertilized plots seeded by JD, spring barley produced higher yields than winter and spring wheat (Tables 9, 10). Winter and spring wheat bundle yields were not significantly different although spring combine yields were 75 percent of winter wheat combine yields. In fertilized plots seeded by CP, spring barley produced higher yield than winter wheat. In unfertilized plots seeded with JD, the yield of spring barley was significantly higher than that of winter wheat and spring wheat. Spring wheat yield was higher than winter wheat yield. Combine yields followed the same trend (Table 9). The yield of unfertilized plots of winter wheat and spring barley seeded by CP were not significantly different. In general, there were no significant differences in bundle yields of either winter wheat or spring barley planted by either drill. In 2003 only the CP was used. Yields were significantly different between the fertilized plots of winter wheat, spring barley and spring wheat (Tables 11, 12). Spring barley produced the highest yields followed by winter wheat and spring wheat. The high yield in barley was probably attributed to a high number of heads/ft2 (Table 11).
Conventional tillage vs. no-till
For both 2002 and 2003, the yield of the fertilized conventional and no-till winter wheat was not significantly different. For the unfertilized plots, the yield of no-till winter wheat was significantly lower than that of conventional winter wheat in both years (Tables 9, 10, 11, 12). In fertilized plots for both years, conventional spring wheat yields were significantly higher than yields of no-till spring wheat yields. The results were similar for unfertilized plots. In fertilized spring barley, conventional plots produced significantly higher yields than no-till plots in 2002 but the differences were not significant in 2003 (Tables 9, 10, 11, 12). In unfertilized plots, conventional spring barley plots produced significantly higher yields than no-till spring barley plots seeded by CP in 2002. No significant differences were observed in JD seeded plots. In 2003, for unfertilized plots, conventional spring barley produced significantly higher yield than no-till spring barley plots (Tables 9, 10, 11, 12).
Summary
In both conventional and no-till cropping systems, spring barley produced the highest yield followed by winter wheat and then spring wheat. In winter wheat cropping systems, there was no difference in yield between conventional and no-till systems. Spring wheat and spring barley produced higher yields under conventional tillage than under no-till cropping system.
Crop residue management
No plant counts were done in this experiment in 2002. In 2003 plant stand ranged from 73 to 99percent of target stand (Table 11). The lowest stands were observed in check plots (no N) and the highest stands were observed in manure- and N-applied plots. Grain yields were generally higher in 2003 than in 2002, probably due to more precipitation received in 2003. In 2002, grain yield from the manure treatment was significantly higher than other treatments except the pea vine treatment (Table 9). Grain yields of all the N-fertilized treatments, with or without spring burning, were not significantly different. Grain yields of all the unfertilized plots, with spring or fall burn, were not significantly different and were the lowest yields. Grain yield was highly correlated with heads/ft2 (r = 0.71, P < 0.0001) and the manure treatment had significantly higher heads/ft2 than other treatments (Table 9). The heads/ft2 of the pea vine treatment ranked second. In 2003, the manure treatment produced the highest grain yield although this yield was not significantly different from the 40-N and 80-N treatments (Table 11). The lowest yields were obtained on the unfertilized spring and fall burn treatments. Grain yield was highly correlated with heads/ft2 (r = 0.62, P < 0.0001) and test weight (r = 0.86, P < 0.0001). These components were high in the manure and N treatments (Table 10). Data on how much N was added by manure and pea vine are not yet available.
Summary
High yields were obtained when manure or N was applied. Field burning without N application resulted in the lowest yields.
Wheat/pea
In 2002 the pea plant stand was >80 percent of the target stand (Table 9). In 2003, the plant stand was >80 percent except for the no-till pea plots where only 66 percent of the target stand was observed (Tables 9, 11). No wheat stands were determined in 2002. In 2003, wheat stands were >80 percent of target stand except for the no-till plots where the stand was 75 percent of the target stand (Table 11).
Under the wheat system, tillage treatments influenced grain yield in both 2002 and 2003 (Tables 9, 11). In 2002, highest grain yields were produced when plots were plowed in the fall or spring. Grain yield was correlated with bundle yield (r = 0.81, P < 0.0001) and therefore yield components obtained from bundle samples can be used to explain yield variations in combine data. Grain yield was correlated to heads/ft2 (r = 0.66) and plots plowed in the fall and spring had high numbers of heads/ft2. Data on weed infestation were not collected in 2002. In 2003, the highest yield was obtained from plots plowed in the spring and fall (Table 11). The lowest yield was produced under no-till where downy brome infestation was significantly higher compared to other treatments. No significant differences were observed among treatments on heads/ft2 and kernel weight (Table 11).
In 2002, tillage treatments under the pea system significantly influenced grain yield. The no-till system produced the highest pea yields and plowing in the fall resulted in the lowest pea yield (Table 9). In 2003, there were no significant differences in grain yield among all the tillage treatments. However, the lowest yield was produced when plots were plowed in the fall (Table 11). Plant stand did not influence yield (Tables 9, 11).
Summary
For wheat, highest yields were produced when plots were plowed in the fall or spring. The lowest yield was produced in the no-till system where downy brome infestation was significantly high. For peas, the highest yield was produced under the no-till system and the lowest yield was produced when plots were plowed in the fall.
Tillage fertility
Data from the tillage fertility experiment are obtained in alternate years. No data were available in 2002 and data reported below are from 2003. Plant stands in the tillage-fertility plots were not significantly different from each other and ranged from 81 to 84 percent of the target stand (Table 11).
There was no significant interaction between tillage and fertility treatments. Increasing fertilizer rates from 0 to 120 lb N/acre significantly increased grain yield of wheat from 40 to 62 bu/acre (Table 11). Increasing the rate to 160 lb N/acre significantly depressed yields to 59 bu/acre. Grain yield was correlated to heads/ft2 (r = 0.56, P < 0.0001). Head counts increased with increasing N rates. Plant stand was not significantly different among the treatments (Table 12) indicating that increased N rates increased tillering. Grain yield at the highest N rate was reduced, probably because soil moisture was depleted before maturity due to increased water demand imposed by increased tillering. The low test weight at the highest N rate is indicative of drought stress (Table 10). The 2003 spring was drier and warmer than normal (Table 8). Downy brome counts were not significantly different among the N treatments (Table 11). Tillage treatments influenced grain yield. The grain yield of the plowed plots was significantly higher than the yield of disk and sweep tillage plots (Table 11). The plow treatment had the highest heads/ft2 and almost no downy brome infestation compared to the other treatments (Table 11). The sweep treatment, which produced the lowest yield, had the highest downy brome infestation (Table 11).
Summary
Increasing N rates up to 120 lbs/acre increased yield through high numbers of heads/ft2. The highest yield was produced under the plowed plots, which had a higher numbers of heads/ft2 than the disked and swept plots. The sweep treatment produced the lowest yield, probably because of a high downy brome infestation.
No-till wheat (summer fallow)
In both no-till treatments, in 2002 and 2003, grain yields generally increased with increasing N fertilization up to 80 lbs/acre, then slightly declined (Table 13). There was little difference between the older set of plots (A) and the newer set of plots (B). Grain yields in the conventional tillage plots, with the same N fertilization rates, were consistently greater. Residue yields showed similar trends as the grain yields (Table 13). Test weights were similar among treatments; however they were generally greater in 2003 than 2002 (Table 13). The 1,000-kernel weights decreased with increasing N fertilization, but were similar between N treatments for both years regardless of the age of the treatments (Table 13).
[an error occurred while processing this directive]