http://orcid.org/0000-0001-6273-2301
Figures
Abstract
A two-year rotation of summer fallow (SF)/winter wheat (WW) is the most common cropping system in low precipitation areas of the U.S. Pacific Northwest (PNW). In SF, multiple tillage operations are used to manage weeds and maximize soil water storage and potential WW yield. Reduced tillage fallow (RTF) is an alternative to SF that leaves >30% of the previous crop’s residue on the surface. A four-year (2014–18) field study was conducted to evaluate the influence of SF and RTF on weed species density, cover and composition in dryland WW; determine if changes in these weed infestation attributes have any influence on crop density and yield; and evaluate economic costs of each type of fallow management. The experimental design was randomized complete block with four replications where each phase of SF/WW and RTF/WW rotations was present every year. Individual plots of WW were divided into a weedy sub-plot with no weed control, general area with chemical weed control, and weed-free sub-plot where weeds were manually removed. Infestations of annual grass and other weeds in weedy sub-plots increased throughout the study. Grass weed cover, consisting mainly of downy brome (Bromus tectorum L.), and total weed cover were significantly lower in WW following RTF than following SF in all years except 2018. Densities of grass and total weeds were similar in both fallow managements indicating that weed plants were larger in WW following SF than following RTF due to earlier or faster emergence. Grass cover differences were not found in general areas likely because of a reduced seedbank. When weeds were present, mean yield of WW was higher following RTF than SF indicating that weeds were less competitive in RTF. Reduced tillage fallow could improve weed management in fallow/WW cropping systems of the PNW compared to SF/WW, particularly if the most problematic species are grasses.
Citation: San Martín C, Long DS, Gourlie JA, Barroso J (2018) Weed responses to fallow management in Pacific Northwest dryland cropping systems. PLoS ONE 13(9): e0204200. https://doi.org/10.1371/journal.pone.0204200
Editor: Jose Luis Gonzalez-Andujar, Instituto Agricultura Sostenible, SPAIN
Received: February 28, 2018; Accepted: September 5, 2018; Published: September 20, 2018
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: Data are available in a csv file as supporting information.
Funding: This work was funded from USDA Agricultural Research Service National Program in Agricultural Systems Competitiveness and Sustainability (NP 216), to D.S.L. This work was also supported by a grant award from the Oregon Wheat Commission (http://www.owgl.org/) to J.B. The Oregon Wheat Commission had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. USDA Agricultural Research Service had a role in the decision to publish the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The most common cropping system in the winter wheat (WW, Triticum aestivum L.) producing region of the U.S. Pacific Northwest is WW followed by one year of fallow [1,2]. The fallow year is used to recharge soil moisture [3,4] as needed to maximize potential WW yield in the following year [5,6]. Conventional summer fallow (SF) involves the use of frequent tillage (e.g., disk harrowing, rod weeding [7]) to control weeds and create a dust mulch that resists evaporation [3,7,8]. Unfortunately, SF is prone to wind erosion that reduces soil productivity and produces dust emissions that impairs air quality and threatens human health [4,9]. To allay concerns about wind erosion, growers are interested in adopting conservation tillage practices such as minimum tillage and no-till [1]. Conservation tillage can be less expensive than SF because of savings of fuel, machinery, labor and time [10,11]. However, high herbicide costs in some years may favor SF [11].
No-till practices involving chemical fallow (CF, controlling weeds in fallow fields with chemicals) have been used successfully in the Great Plains where a summer precipitation pattern helps sustain crops through the hot summer season [12,13]. However, summers are hot and dry in the semiarid PNW and the absence of tillage that otherwise would interrupt capillary flow may cause severe evaporative loss of soil moisture [14,15]. In general, WW yields tend to be higher in rotation with SF than with CF [6,8] because growers can seed WW earlier into SF than into CF [16,17].
Conservation tillage practices minimize soil disturbance, leave vegetative residue on the surface, and rotate primary crop with other crops [18]. Reduced tillage fallow (RTF) involves a single tillage operation with a minimum disturbance implement such as V-blade sweep. Wind erosion is reduced by retaining surface residue and soil clods that increase surface roughness [4,19]. Schillinger [1] showed that crop yield increased following RTF compared to CF. Cardina et al. [20] found that CF increased weed seedbank density, due to lack of seed burial, compared with tillage-based systems.
Changes in tillage regime may cause floristic inversion due to changes in the seedbank [21–23]. Incorporating seeds deeper into the soil with tillage might favor conditions for an increase of seedbank [23,24], but burying seeds might also avoid seed germination [25] and favor seed deterioration [26]. Under no-till and RTF, seeds are more likely to remain near the soil surface [24] where they are susceptible to insect and bird predation in the summer periods [27]. Germination of some species is favored when seeds are near the soil surface and thus no-till and RTF will increase seedling emergence compared with conventional tillage. However, Mohler [24] observed that emergence increased in the first year followed by reduction in later years due to depletion of the seed surface fraction by the action of herbicides and shallow cultivation. Moreover, crop residues associated with no-till and RTF may suppress weeds by blocking sunlight and reducing physical space for seedling emergence [28–30].
Common weeds in the WW phase of the fallow / WW system are winter annual grasses: downy brome and feral rye (Secale cereale L.); winter annual broadleaves: prickly lettuce (Lactuca serriola L.), tumble mustard (Sisymbrium altissimum L.), and flixweed (Descurainia sophia (L.) Webb. Ex Prantl); and summer annual broadleaves: Russian thistle (Salsola tragus L.), lambsquarters (Chenopodium ssp.), tumble pigweed (Amaranthus albus L.), fiddleneck (Amsinckia intermedia Fisch. & C.A. Mey), horseweed (Conyza canadiensis (L.) Cronq.), and prostrate knotweed (Polygonum aviculare L) [3]. Horseweed and prickly lettuce have increased in CF in the region [3]. Lambsquarters has increased in tilled systems more than in no-till [20]. Russian thistle has decreased in no-till systems [3]. In general, annual wind-disseminated weeds and grasses tend to occur in reduced and no-till systems whereas non-wind-disseminated weeds are more frequent in tillage systems [31,32]. No-till may favor weeds whose seeds concentrate in the soil surface and are normally not a problem in tillage systems [33,34].
Downy brome, which is one of the most problematic species in WW in the PNW region [3], germinates mainly in fall with first rains and to a lesser extent in early spring [35] to take advantage of moisture and nutrients in advance of crop growth. Similar morphology (life cycle mimicry) to wheat and limited chemical control options help proliferate this weed. In eastern Washington, downy brome infestations of 54 and 500 plants m-2 caused 28% to 100% reduction in WW yield [36]. Lyon and Baltensperger [37] found that two-year crop rotations were effective in controlling downy brome if plants were not allowed to produce seeds in the fallow periods. Wicks [38] showed that moldboard plowing effectively controlled this weed species during fallow owing to deep seed burial. Downy brome control might be improved in no-till cropping systems with the help of occasional tillage [39,40].
Fallow management is an important determinant of weed infestations in the following cropping year. Therefore, amount of crop residue and degree of soil disturbance will influence the composition and density of weeds. The objectives of this research were to: (1) evaluate the influence of SF and RTF on density, cover and composition of downy brome and other weed species in dryland WW; (2) determine if changes in these weed infestation attributes influence crop density and yield; and (3) evaluate economic costs of each type of fallow management.
Materials and methods
Site description and history
The study was undertaken at a 6-ha site near Echo, OR, USA (45.7207o N. Lat., -119.0483o W. Lon.). The soil is a well-drained Ritzville silt loam (coarse-silty, mixed, superactive, mesic Calcidic Haploxeroll) with a surface soil pH of 5.9. Monthly mean temperature was computed from hourly measurements collected at a standard weather station located at the study site (Fig 1). Daily measurements of precipitation were obtained with a tipping bucket rain gauge at the site and used to compile the total precipitation for each month. Mean annual precipitation is 269.4 mm based on long-term records obtained at this site since 2001.
a) 2014–2015, b) 2015–2016, c) 2016–2017, and d) 2017–2018. The bar graph indicates accumulated rainfalls per month (mm), and the line the average temperature per month (°C). *Precipitation gage was not working properly; this value is a distance-weighted average of nearby Bureau of Reclamation AgriMet Echo site (45.7186o N. Lat., 119.3111 o W. Lon.) and Citizens Weather Observer Program DW8844 site (45.7255o N. Lat., 118.9250 o W. Lon.).
Experimental design
The experiment was conducted over four growing seasons (2014–2015, 2015–2016, 2016–2017, and 2017–2018). Two cropping systems or rotations, SF/WW and RTF/WW, were established in 2014. The experimental design was a randomized complete block with four replications. Each phase of the rotation (i.e., crop and fallow) was present every season. Therefore, each block had four plots consisting of two WW plots, a fallow plot managed with tillage (SF), and a fallow plot managed with reduced tillage (RTF) bringing the total number of plots to 16 per season. Individual plots were 30.5 m × 6.1 m (100 ft × 20 ft).
Soft white winter wheat (cv. Bobtail in 2015, 2017 and 2018, and cv. Ovation in 2016) was sown into plots with two passes of a 3-m wide Seed Hawk (Langbank, SK, Canada) air till drill with hoe-type openers on 30 cm row spacing. Different crop varieties were used to achieve the highest yield that might be possible for the anticipated conditions as a regular practice in the region (Ovation offers stripe rust tolerance whereas Bobtail offers foot rot resistance). Sowing was performed on 30 November 2014 (delayed due to lack of soil moisture), 13 October 2015, 6 October 2016, and 6 October 2017. Seed density was 132 kg ha-1 in 2015, 127 kg ha-1 in 2016, 119 kg ha-1 in 2017, and 99 kg ha-1 in 2018. Fertilizer was applied each year as a solution of 79% urea-ammonium nitrate and 21% ammonium thiosulfate to give rates of 55 kg N ha-1 and 11 kg S ha-1 (0.794 kg of product in 166.6 l ha-1). Herbicide applications (Table 1) were performed with a self-propelled sprayer delivering 187.2 l ha-1. Harvest occurred at physiological maturity in mid-July using a small plot combine harvester equipped with 1.9 m header and 9–10 mm top sieve.
Winter wheat plots were subdivided to two small sub-plots: a weedy sub-plot (6.1 m × 3.05 m) where no herbicide was applied and a weed-free sub-plot (6.1 m × 3.05 m) where weeds were pulled by hand. The remaining portion of the plot (6.1 m × 24.4 m) was reserved for general weed management with post-emergence herbicides commonly used in the region.
Fallow management
The field was in WW in 2012 and it was fallowed in 2013. Starting in 2014, each phase of each rotation was present each year. Winter wheat plots harvested in 2014 completed three cycles of the rotation by 2018 whereas those harvested in 2015 completed two cycles by 2018. Though WW scheduled for harvest in 2015 by design followed fallow, it actually followed two years of fallow because of fallowing of the field in 2013 and presense of a fallow phase in 2014. Weed management in SF consisted of one pass with a disk harrow at a soil depth of 13–15 cm in early spring followed by several passes of a rotary rod weeder in late spring and summer at a depth of 9 cm (Table 2). In RTF, tillage consisted of one pass with an undercutter V-blade sweep in early summer at a depth of 13 cm. Weed control in RTF during spring and summer also included applications of glyphosate (N-(phosphonomethyl) glycine) and other herbicides as needed (Table 2). After harvest, all plots were treated with paraquat (N,N′-dimethyl-4,4′-bipyridinium dichloride) to control late-season broadleaf weeds that re-grow. Herbicides were applied using a self-propelled sprayer with 6-m spray boom delivering 93.6 l ha-1.
Data collection
Weed density and cover were measured within 16 sampling frames (1 m × 0.5 m) placed in each WW plot: four in the weedy sub-plot, four in the weed-free sub-plot, and eight in the general area. Sampling frames were constructed from 1.25 cm diameter PVC pipe. Winter wheat plots were sampled three times during the growing season: early season at the beginning of weed competition (mid-March, tillering), mid-season at peak crop growth (end of May, flowering), and late-season at crop maturity (end of June, near harvest). At each sampling time, percent cover of the crop and each weed species were estimated visually within each frame. Density (plants m-2) of the crop and each weed species were determined by counting the number of plants in each frame. At harvest (i.e., late season sampling), the standing wheat in each frame was collected by destructive sampling and bundled. Bundles of wheat were then threshed, and the resulting grain was cleaned and weighed for determination of yield.
Statistical analysis
A generalized linear mixed model (GLMM) was fit to test for differences between SF and RTF in terms of density and cover of grass weeds, broadleaf weeds, and total weeds. Models were fit with different data distributions by year and plot area (weedy vs. general) because weed pressure significantly differed among years and between plot areas each growing season. Mixed models allowed for incorporation of different levels of pseudo-replication inherent to plot sampling. Type of fallow management, crop density or crop cover, and their interactions were entered as fixed effects. Crop density and cover were centered around their mean. Plot was included as a random effect. In all models, assumptions of equal variances and normal distribution of residuals were evaluated graphically. These analyses were implemented with thelme4 [41] package in the R program v. 3.3.2 [42].
Non-linear regression was used to test the effect of weeds on wheat yield in relation to fallow management. The non-linear models were the hyperbolic rectangular [Eq 1] [43,44] and negative exponential [Eq 2] [45,46]:
[Eq 1]
[Eq 2]
where Y is yield in presence of weeds (kg ha-1), Yo is yield in absence of weeds (kg ha-1), X is either weed density (plants m-2) or weed cover (%), a is yield loss per X unit as weed presence approaches zero, b is yield loss as X approaches infinity, and c estimates the rate of reduction in crop yield as weed density or cover increases. Models were selected based on the corrected Akaike information criterion (AICc). The non-linear mixed model (NLMM) analysis was used to identify differences between model parameters (Yo and c). These parameters were included as fixed effects and plot as a random effect. These analyses were implemented with the nlme [47] package in the R program [42].
Those frames in which a species is present vs. total number of frames was calculated to determine percent species presence in each year. Non-parametric multivariate analysis of variance (PERMANOVA) was conducted with “adonis” function available in vegan package [48] of the R program [42] to search for significant differences between the weed community of the two cropping systems. Type of fallow management, replicate, and their interaction were used as factors.
Cost comparison
The total U.S. dollar cost of each fallow management was estimated by year, by summing the individual costs of the different management operations (Table 2). Cost of various mechanical and chemical operations in fallow/WW production systems of the area are quoted in Seavert et al. [49] at $11.10 ha-1 for herbicide application, $24.70 ha-1 for disk harrowing, $14.80 ha-1 for rod weeding, and $14.80 ha-1 for sweep cultivation. Each herbicide application cost per hectare was calculated by multiplying the application rate times the retail prices for individual herbicides listed online at the University of Nebraska [50]. For example, if application rate is 4.67 L ha-1 (3081 g ai ha-1) and retail price of Roundup Power Max® is $6.60 L-1 (660 g a.i.), then herbicide application cost is $30.80 ha-1.
Results and discussion
Weed composition and year effect
Total precipitation during each cropping year (September to August) was 273 mm (101% of long-term average) in 2015, 262 mm (97%) in 2016, 327 mm (121%) in 2017, and 246.2 mm (92%) in 2018. Mean density (plants m-2) of total weeds in weedy sub-plots increased from 7.9 ± 5.8 in 2015 to 25.4 ± 24.0 in 2016, 51.6 ± 32.4 in 2017, and 88.0 ± 44.2 in 2018. Mean cover (%) of total weeds also increased with year from 21.0 ± 23.4 in 2015 to 26.1 ± 28.8 in 2016, 79.0 ± 44.7 in 2017, and 75.8 ± 33.3 in 2018. Dominant weeds were tumble mustard, downy brome, prickly lettuce, Russian thistle, and volunteer wheat as indicated by percentage of presence (Table 3). In eastern Washington, Schillinger [1] found downy brome and Russian thistle to be the most dominant species whereas tumble mustard had a minor presence. These weed species were influenced by growing conditions. Other studies have found that the amount and distribution of precipitation strongly influenced the seed germination [51]. In this study, downy brome, tumble mustard and prickly lettuce increased over time while Russian thistle and kochia decreased. More species appeared during our study (e.g., coast fiddleneck, flixweed, panicle willow weed and horseweed) than disappeared (e.g., lambsquarters). Significant differences in weed community composition between SF/WW and RTF/WW were not found by multivariate analyses.
Grass weed infestation was lower in 2015 than 2016, 2017 and 2018 apparently due to seedbank depletion over two years of fallow in which a single application of glyphosate was applied each year. Downy brome increased the most from the beginning of the study in terms of species presence (percentage of frames in which a species is present vs. total number of frames) and abundance (plants m-2 and/or percentage cover). Downy brome abundance may have increased because of the absence of a post-emergence herbicide application on WW plots in 2015 due to the low density of this species and other grass weeds in that year (0.56 plants m-1). However, increased precipitation, especially in 2017, likely enhanced conditions for germination and development of downy brome resulting in its increased presence and abundance in that year. Due to the great proliferation of downy brome in 2017, the seedbank could have increased, and consequently led to a great abundance in the following growing season in 2018. Schillinger [1] and Blackshaw et al. [51] found that favorable moisture conditions increased density of downy brome and flixweed. Flixweed is a species with a similar biology to tumble mustard that also increased in our study under those conditions. In contrast, Russian thistle infestation decreased over the three years of this study. Drier weather conditions and the absence of other weeds may have favored this species in 2015. Russian thistle was found to have a low tolerance to competition, growing better in crop- and weed-free sub-plots [52] as well as after harvest. Schillinger [1] also reported scarcity of this species in humid years, as did Blackshaw et al. [51] who found more density of this species in low rainfall periods. Year 2015 was the warmest of the four growing seasons, which could explain the greatest Russian thistle presence in that year along with the late spring rainfall (Fig 1) that could have also caused new flushes.
Effect of fallow management on following season’s weeds
For total weeds and broadleaf weeds, differences in density were not detected between fallow managements, except in 2018, where higher broadleaf weed density was obtained in the WW plots managed with RTF. The percentage cover of broadleaf and total weeds was reduced in the WW with RTF compared to SF in 2016 (Table 4, Fig 2), was not affected with fallow management in 2016 and 2017, and increased in 2018 in the RTF compared to SF, although only in the general area. Broadleaf weeds contributed the most to the total weed density (90.3% in 2015, 90.2% in 2016, 72.4% in 2017, and 70.3% in 2018) and cover (94.9% in 2015, 87.0% in 2016, 80.7% in 2017, and 78.3% in 2018). Tumble mustard contributed the most to total broadleaf weed cover and density (56.1% in 2015, 82.8% in 2016, 80.3% in 2017, and 84.9% in 2018). Evans and Young [53] found emergence of tumble mustard was favored when the soil surface is rough and pitted. Though the soil surface in RTF was rougher and with more pits than in SF, tumble mustard density was not significantly different between fallow managements in our study most of the years except for 2018. Prickly lettuce and volunteer wheat did not significantly differ in cover and density between fallow managements (data not shown). Differences in Russian thistle density and cover could not be detected between SF and RTF in 2015 and 2016 (its presence was almost zero in 2017 and 2018). Young and Thorne [3] found that Russian thistle was stimulated by tillage. However, Schillinger [1, 54] did not find differences between tillage regimes and rotations for that species.
Total weed density and cover (a) and grass weed density and cover (b) observed in the weedy and general areas in the cropping systems (Winter wheat (WW)–Summer Fallow (SF) and WW–Reduced Tillage Fallow (RTF)) along the study (2015–2018).
Note: analysis for grasses in 2015 was not included because grasses were almost absent in 2015.
The paucity of weeds in 2015 prevented statistical comparison of grass weed cover between weedy sub-plots of SF and RTF in that year. In contrast, grass weed cover of weedy sub-plots was significantly higher in SF than RTF in 2016 and 2017 (Table 4, Fig 2). The difference in grass weed cover in 2016 may be related to reduced grass weed density in RTF versus SF that year, though a significant difference in grass weed density was not found in 2017. Sampling of weeds was performed at the same time in SF and RTF. Consequently, less grass weed cover at a similar density implies delay in emergence, development, or both in WW following RTF. Therefore, untreated weedy sub-plots had grass infestations that were more competitive in WW plots preceded by SF than WW plots preceded by RTF. In general areas, having a small seedbank due to weed control every year, differences of marginal significance (P < 0.1) were only found in 2016 (Table 4).
Downy brome contributed the most to total grass weed density (70.4% in 2015, 52.5% in 2016, 94.3% in 2017, and 92.4% in 2018). Previous workers reported that emergence of downy brome is favored by uncompacted soils [55,56] in agreement with this study. Summer fallowing by means of disk harrowing and rod weeding creates a finely textured layer of soil with reduced compaction compared to RTF involving a single V-blade sweep operation. Thorne et al. [39] and Kettler et al. [40] found downy brome control improved in no-till cropping systems with occasional tillage. In this study, reducing tillage from two to three operations per season (SF) to one tillage operation per season (RTF) reduced downy brome the following season.
Moreover, leaving higher amounts of residue on the soil surface are known to increase the suppression of weeds [28–30]. Reduced tillage fallow resulted in more crop residue cover and surface roughness than SF, which could explain the higher weed cover found in WW following the SF. Residue suppresses or delays weed germination and emergence by reducing soil temperature [57] and light [29], emitting allelopathic substances [58], and increasing competition for free-space [29,59] but the contribution of each factor to the weed suppression is difficult to determine [59].
However, seed germination inhibited by light exposure (negative photoblastism) has been determined in Bromus species [60]. Del Monte and Dorado [61] found that ripgut brome (Bromus diandrus Roth) seeds buried or shaded superficially can germinate during fall. Recasens et al. [62] found that germination of ripgut brome was favored by chisel plowing compared to moldboard plowing that buried seeds deeper. Intensive tillage in the SF apparently may have buried seeds and limited their exposure to light, which enhanced germination of downy brome in the following crop. Though the shading effect of crop residue might be greater than seed burial, it can suppress downy brome germination and emergence because it works as a physical barrier that can help with weed control. In RTF, the amount of remained crop residue before seeding is larger than in SF and might have contributed to reduce weed development in the WW. However, non-significant differences in weed infestation between fallow managements in 2017, except for the grass cover in the weedy sub-plots, indicates that when the soil moisture is abundant, the effect of residue on weed infestations could be overcome, especially for broadleaf weeds. The same situation could have happened with grasses in 2018 where no differences were found in relation to density and cover for both weedy and general areas (Table 4, Fig 2). In 2018, unusually warm January temperatures coupled with adequate precipitation (Fig 1) might have favored higher, later downy brome germination than in previous years such that fallow management was ineffective and unnoticeable in a well-established crop.
Crop yield in relation to fallow management and weed infestation
Overall, WW yield in the general area was greater in 2015 (3011 kg ha-1) and 2017 (3554 kg ha-1) than in 2016 (1854 kg ha-1) and 2018 (1977 kg ha-1) regardless of fallow management. In 2015, crop yield was not significantly impacted by weed presence due to low weed infestation (5.5 plants m-2). Despite similar amounts of precipitation, 2015 yields were higher than 2016 yields likely due to soil water storage over two years of preceding fallow but also higher May precipitation (Fig 1A). Mean weed infestation in the general area fell from 45 plants m-2 at the beginning of growing season to 11 plants m-2 at harvest. Infestations were not correlated with crop yield in the general area (data not shown). Yield was not impacted by initial weed infestation indicating that weeds were adequately controlled with herbicides.
Mean grain yield in the weedy sub-plots of WW plots was significantly greater in 2015 (2858 kg ha-1) than in 2016 (1360 kg ha-1), 2017 (904 kg ha-1) and 2018 (188 kg ha-1). Yields of weedy sub-plots were only 25.4% and 9.5% of the general area due to high average weed densities of 52 and 88 plants m-2 in 2017 and 2018. From 2016 to 2018, with higher weed pressure than in 2015, grain yield decreased with increasing weed density or cover in the weedy sub-plots (Fig 3). The exponential model (Fig 3) was chosen to represent the relationship between crop yield and weed infestation because it had lower AICc than the hyperbolic model. Parameters Y0 and c in the selected models were non-significant for fallow managements in all years (Table 5). Nevertheless, the curvature of the fit (c) representing the rate of change in grain yield for weed density and weed cover was lower in RTF than in SF, especially for 2017 and 2018, indicating that the negative effect of weeds on crop yield was greater in SF than in RTF. Differences in crop yield between fallow managements were smaller in 2016 than in 2017 and 2018 (Table 5). Crop yield did not differ between fallow management when weed presence was zero (i.e., Y0) whereas the same weed infestations caused larger yield loss in cropping systems with SF than in cropping systems with RTF.
Relationship between crop yield (kg ha-1) and weed density (plants m-2) and weed cover (%) in 2017 (a) and 2018 (b).
Weeds were less competitive in RTF in 2016, 2017 and 2018 resulting in higher mean crop density and cover in the WW following RTF than in the WW following SF in those years. In 2016, the WW following RTF had 200.8 wheat plants m-2 and 35.0% cover compared to 135.3 wheat plants m-2 and 27.1% cover for the WW following SF. In 2017, weed density and cover were 135.8 plants m-2 and 23.0% for WW following RTF, and 84.0 plants m-2 and 16.9% for WW following SF. However, differences were not significant for crop cover in 2017. And in 2018, weed density and cover were 102.8 plants m-2 and 29.1% for WW following RTF, and 60.2 plants m-2 and 21.6% for WW following SF.
Implications for management and concluding remarks
In this study, fallow management influenced weed cover and density in WW, with RTF resulting in the development of smaller plants early in the season and improving suppression of downy brome over that of SF. However, environmental conditions could diminish the effect of RTF as in January 2018 when warm and wet conditions promoted late downy brome germination. Young and Schillinger [63] reported that farmers in the PNW used nearly equivalent amounts of glyphosate and obtained similar profitability under SF and RTF. In this study, weed management costs were similar between SF and RTF in 2016 ($157 vs. $164 ha-1) and 2017 ($181 vs. $195 ha-1) but were larger for RTF in 2015 ($130 vs. $189 ha-1) and 2018 ($107 vs. $175 ha-1). Adjuvants are normally used in commercial fields but in this study they were required in higher amounts to prevent herbicide drift between narrow plots. Therefore, costs obtained for RTF should be considered illustrative. Chemical costs could also be reduced in RTF by rotating herbicides instead of relying upon tank mixtures.
The question remains whether greater use of herbicides in RTF negatively impacts the environment and human health. Reduction in herbicide use might be achieved after several seasons given that the first years after the adoption of conservation tillage, herbicides play an important role in controlling weeds [18]. Sebastian et al. [64] found the seedbank of downy brome became depleted after 4–5 consecutive years of glyphosate applications. A reduction in herbicide application could also be achieved by diversifying the fallow / WW rotation with crops that can break pest cycles. Finally, long-term approaches to integrated weed management should reduce reliance upon herbicides [65] and the extent of herbicide-resistant species [66].
Acknowledgments
The authors thank John McCallum for helpful criticism that improved the manuscript, Steve Umbarger for field operations, and all summer students who helped in field data collection.
References
- 1. Schillinger WF. Seven rainfed wheat rotation systems in a drought-prone Mediterranean climate. F Crop Res. Elsevier B.V.; 2016;191: 123–130.
- 2.
Smiley RW, Siemens MC, Gohlke TM, Poore JK. Small grain acreage and management trends for eastern Oregon and Washington. 2005 Dryland Agricultural Research Annual Report. Pendleton, OR; 2005.
- 3. Young FL, Thorne ME. Weed-species dynamics and management in no-till and reduced-till fallow cropping systems for the semi-arid agricultural region of the Pacific Northwest, USA. Crop Prot. 2004;23: 1097–1110.
- 4.
Papendick RI. Farming with the Wind II: Wind erosion and air quality control on the Columbia Plateau and Columbia Basin. College of Agriculture, Human, and Natural Resource Sci. Spec. Rep. XB1042. Washington State University. Pullman; 2004.
- 5. Wuest SB, Schillinger WF. Evaporation from high residue no-till versus tilled fallow in a dry summer climate. Soil Sci Soc Am J. 2011;75: 1513–1519.
- 6. Schillinger WF, Young DL. Wheat—cropping systems research in the world ‘s driest rainfed wheat region. Agron J. 2004;96: 1182–1187.
- 7. Schillinger WF, Papendick RI. Then and now: 125 years of dryland wheat farming in the Inland Pacific Northwest. Agron J. 2008;100: 166–182.
- 8. Juergens LA, Young DL, Schillinger WF, Hinman HR. Economics of alternative no-till spring crop rotations in Washington’s wheat–fallow region. Agron J. 2004;96: 154–158.
- 9. Upadhyay BM, Young DL, Wang HH, Wandschneider P. How do farmers who adopt multiple conservation practices differ from their neighbors? Am J Altern Agric. 2003;18: 27–36.
- 10. Hobbs PR. Conservation agriculture: What is it and why is it important for future sustainable food production? J Agric Sci. 2007;145: 127–137.
- 11. Gray RS, Taylor JS, Brown WJ. Economic factors contributing to the adoption of reduced tillage technologies in central Saskatchewan. Can J For Res. 1996;76: 661–668.
- 12. Aase JK, Pikul JL. Crop and soil response to long-term tillage practices in the northern Great Plains. Agron J. 1995;87: 652–656.
- 13. Smika DE. Fallow management practices for wheat production in the central Great Plains. Agron J. 1990;82: 319–323.
- 14. Hammel J, Papendick R, Campbell G. Fallow tillage effects on evaporation and seed zone water content in a dry summer climate. Soil Sci Soc Am J. 1981;45: 1016–1022.
- 15. Wuest SB. Tillage depth and timing effects on soil water profiles in two semiarid soils. Soil Sci Soc Am J. 2010;74: 1701–1711.
- 16. Higginbotham RW, Jones SS, Carter AH. Adaptability of wheat cultivars to a late-planted no-till fallow production system. Sustainability. 2011;3: 1224–1233.
- 17. Donaldson E, Schillinger WF, Dofing SM. Straw production and grain yield relationships in winter wheat. Crop Sci. 2001;41: 100–106.
- 18. Rahman M. Weed management in conservation agriculture. Adv Plants Agric Res. 2017;7: 253.
- 19. Thorne ME, Young FL, Pan WL, Bafus R, Alldredge JR. No-till spring cereal cropping systems reduce wind erosion susceptibility in the wheat/fallow region of the Pacific Northwest. J Soil Water Conserv. 2003;58: 250–257.
- 20. Cardina J, Herms CP, Doohan DJ, Cardina J, Doohan DJ. Crop rotation and tillage system effects on weed seedbanks. Weed Sci. 2002;50: 448–460.
- 21. Ball DA. Weed seedbank response to tillage, herbicides, and crop rotation sequence. Weed Sci. 1992;40: 654–659.
- 22. Dorado J, Del Monte JP, López-Fando C. Weed seedbank response to crop rotation and tillage in semiarid agroecosystems. Weed Sci. 1999;47: 67–73.
- 23. Buhler DD, Hartzler RG, Forcella F. Implications of weed seedbank dynamics to weed management. Weed Sci. 1997;45: 329–336.
- 24. Mohler CL. A model of the effects of tillage on emergence of weed seedlings. Ecol Appl. 1993;3: 53–73. pmid:27759227
- 25. Chauhan BS, Gill G, Preston C. Influence of environmental factors on seed germination and seedling emergence of rigid ryegrass (Lolium rigidum). Weed Sci. 2006;54: 1004–1012.
- 26. Gómez R, Liebman M, Munkvold G. Weed seed decay in conventional and diversified cropping systems. Storkey J, editor. Weed Res. 2013;54: 13–25.
- 27. Baraibar B, Canadell C, Torra J, Royo-Esnal A, Recasens J. Weed seed fate during summer fallow: the importance of seed predation and seed burial. Weed Sci. 2017;65: 1–10.
- 28. Crutchfield DL, Wicks GA, Burnside OC. Effect of winter wheat (Triticum aestivum) straw mulch level on weed control. Weed Sci. 1986;34: 110–114.
- 29. Teasdale JR, Mohler CL. The quantitative relationship between weed emergence and the physical properties of mulches. Weed Sci. 2000;48: 385–392.
- 30. Fernandez R, Quiroga A, Noellemeyer E, Funaro D, Montoya J, Hitzmann B, et al. A study of the effect of the interaction between site-specific conditions, residue cover and weed control on water storage during fallow. Agric Water Manag. 2008;95: 1028–1040.
- 31. Tørresen KS, Skuterud R. Plant protection in spring cereal production with reduced tillage. IV. Changes in the weed flora and weed seedbank. Crop Prot. 2002;21: 179–193.
- 32. Swanton CJ, Shrestha A, Roy RC, Ball-Coelho BR, Knezevic SZ. Effect of tillage systems, N, and cover crop on the composition of weed flora. Weed Sci. 1999;47: 454–461.
- 33. Yenish JP, Doll JD, Buhler DD. Effects of tillage on vertical distribution and viability of weed seed in soil. Weed Sci. 1992;40: 429–433.
- 34. Clements DR, Benoit DL, Murphy SD, Swanton CJ. Tillage effects on weed seed return and seedbank composition. Weed Sci. 1996;44: 314–322.
- 35. Lyon DJ, Hulting AG, Morishhita DW, Young FY. Integrated management of downy brome in winter wheat. Pacific Northwest Ext PNW668. 2015; 1–6.
- 36. Rydrych DJ, Muzik TJ. Downy brome competition and control in dryland wheat. Agron J. 1968;60: 279–280.
- 37. Lyon DJ, Baltensperger DD. Cropping systems control winter annual grass weeds in winter wheat. J Prod Agric. 1995;8: 535–539.
- 38. Wicks GA. Integrated systems for control and management of downy brome (Bromus tectorum) in cropland. Weed Sci. 1984;32: 26–31.
- 39. Thorne ME, Young FL, Yenish JP. Cropping systems alter weed seed banks in Pacific Northwest semi-arid wheat region. Crop Prot. 2007;26: 1121–1134.
- 40. Kettler TA, Lyon DJ, Doran JW, Powers WL, Stroup WW. Soil quality assessment after weed-Control tillage in a no-till wheat–fallow cropping system. Soil Sci Soc Am J. 2000;64: 339–346.
- 41. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67: 1–48.
- 42.
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing [Internet]. 2016. Available: https://www.r-project.org/.
- 43. Cousens RD. A simple model relating yield loss to weed density. Ann Appl Biol. 1985;107: 239–252.
- 44. Canner SR, Wiles LJ, Mcmaster GS. Weed reproduction model parameters may be estimated from crop yield loss data. Weed Sci. 2002;50: 763–772.
- 45. Barroso J, Alcantara C, Saavedra M. Competition between Avena sterilis ssp. sterilis and wheat in South Western Spain. Spanish J Agric Res. 2011;9: 862–872.
- 46. Torner C, González-Andújar J, Fernández-Quintanilla C. Wild oat (Avena sterilis L) competition with winter barley: plant density effects. Weed Res. 1991;31: 301–307.
- 47. Pinheiro J, Bates D, DebRoy S, Sarkar D. nlme: Linear and nonlinear mixed effects models. R package version 3.1–131. 2017.
- 48. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: Community ecology package. R package version 2.4–4 [Internet]. 2017. Available: https://cran.r-project.org/package=vegan
- 49.
Seavert C, Macnab S, Sharp K, Petrie S. Enterprise budget wheat (winter) following fallow, conservation tillage, less than 12-inch precipitation zone, North Central Region. Oregon State University Extention Service. 2012.
- 50.
Anonymous. 2017 Approximate Retail Price ($) per unit of selected herbicides for field crops. Univeristy of Nebraska-Lincoln [Internet]. 2017 [cited 29 Dec 2017]. Available: https://cropwatch.unl.edu/2017-CW-News/2017-documents/weed-management/UNL-EC130-Herbicide-Prices.pdf
- 51. Blackshaw RE, Larney FJ, Lindwall CW, Watson PR, Derksen D a. Tillage intensity and crop rotation affect weed community dynamics in a winter wheat cropping system. Can J Plant Sci. 2001;81: 805–813.
- 52. Young FL. Russian thistle (Salsola iberica) growth and development in wheat (Triticum aestivum). Weed Sci. 1986;34: 901–905.
- 53. Evans RA, Young JA. Microsite requirements for establishment of annual rangeland weeds. Weed Sci. 1972;20: 350–356.
- 54. Schillinger WF. Ecology and control of Russian thistle (Salsola iberica) after spring wheat harvest. Weed Sci. 2007;55: 381–385.
- 55. Thill D., Schirman R., Appleby A. Influence of soil moisture, temperature, and compaction on the germination and emergence of downy brome (Bromus tectorum). Weed Sci. 1979;27: 625–630. Available: http://www.jstor.org/stable/4043082
- 56. Kyle GP, Beard KH, Kulmatiski A. Reduced soil compaction enhanves establishment of non-native plant species. Plant Ecol. 2007;193: 223–232.
- 57. Munawar A, Blevins RL, Saul MR. Tillage and cover crop management for soil water conservation. Agron J. 1990;82: 773–777.
- 58. Alsaadawi IS. Allelopathic influence of decomposing wheat residues in agroecosystems. J Crop Prod. 2001;4: 185–196.
- 59. Teasdale JR. Contribution of cover crops to weed management in sustainable agricultural systems. J Prod Agric. 1996;9: 431–476.
- 60. Jauzein P. Photosensitivity of the annual bromus (Bromus L.). Weed Res. 1989;29: 53–63.
- 61. Del Monte JP, Dorado J. Effects of light conditions and after-ripening time on seed dormancy loss of Bromus diandrus Roth. Weed Res. 2011;51: 581–590.
- 62. Recasens J, García AL, Cantero-Martínez C, Torra J, Royo-Esnal A. Long-term effect of different tillage systems on the emergence and demography of Bromus diandrus in rainfed cereal fields. Weed Res. 2016;56: 31–40.
- 63. Young DL, Schillinger WF. Wheat farmers adopt the undercutter fallow method to reduce wind erosion and sustain profitability. Soil Tillage Res. Elsevier B.V.; 2012;124: 240–244.
- 64. Sebastian DJ, Nissen SJ, Sebastian JR, Beck KG. Seed bank depletion: the key to long-term downy brome (Bromus tectorum L.) management. Rangel Ecol Manag. 2017;70: 477–483.
- 65. Mortensen DA, Egan JF, Maxwell BD, Ryan MR, Smith RG. Navigating a critical juncture for sustainable weed management. Bioscience. 2012;62: 75–84.
- 66. Heap I, Duke SO. Overview of glyphosate-resistant weeds worldwide. Pest Manag Sci. 2018;74: 1040–1049. pmid:29024306