Growing a crop in place of fallow may improve soil properties but result in reduced soil water and crop yields in semiarid regions. This study assessed the effect of replacing fallow in no-till winter wheat (Triticum aestivum L.)–fallow with cover, forage, or grain crops on plant available water (PAW), wheat yield, grain quality, and profitability over 5 yr, from 2007 to 2012. Plant available water at wheat planting was reduced the most when the fallow period was the shortest (i.e., following grain crops) or when biomass production was the greatest. Winter and spring lentil (Lens culinaris Medik.) produced the least biomass, used the least soil water, and had the least negative effect on yield. For every 125 kg ha⁻¹ of cover or forage biomass grown, PAW was reduced by 1 mm, and for every millimeter of PAW, wheat yield was increased by 5.5 kg ha⁻¹. There was no difference in wheat yield whether the preceding crop was harvested for forage or left as standing cover. In years with above-average precipitation, wheat yield was reduced 0 to 34% by growing a crop in place of fallow. However, in years with below-average precipitation, wheat yield was reduced 40 to 70% without fallow. There was minimal negative impact on wheat yield growing a cover or forage crop in place of fallow if wheat yield potential was 3500 kg ha⁻¹ or greater. Net returns were reduced 50 to 100% by growing a cover crop. However, net returns were increased 26 to 240% by growing a forage crop. Integrating annual forages into the fallow period in semiarid regions has the greatest potential for adoption.

Abbreviations

ET: evapotranspiration
PAW: plant available water
W-F: winter wheat–fallow rotation
W-GP: winter wheat–field pea rotation
W-SC-F: winter wheat–summer crop–fallow rotation
W-W: continuous winter wheat
WUE: water use efficiency

Winter wheat (Triticum aestivum L.) is the predominant crop grown in the Central Great Plains. It is commonly grown either in a 2-yr rotation of winter wheat–fallow (W-F) or 3-yr rotation of winter wheat–summer crop (e.g., corn [Zea mays L], grain sorghum [Sorghum bicolor (L.) Moench], or sunflower [Helianthus annuus L.]) –fallow (W-SC-F) (Norwood et al., 1990). Fallow has been a common practice in semiarid regions of the world to store soil water for subsequent crops and stabilize crop yields (Nielsen and Calderón, 2011). The fallow period in W-F lasts ∼15 mo and does not store soil water as efficiently as the 11-mo fallow period in W-SC-F (Norwood et al., 1990; Norwood, 1994), yet many growers who receive <500 mm annual precipitation still commonly use W-F to reduce the risk of low crop yields.

The use of fallow is often at the expense of increased operating costs (Dhuyvetter et al., 1996), soil organic carbon depletion (Peterson et al., 1998; Sherrod et al., 2003; Blanco-Canqui et al., 2010), increased wind and water soil erosion (Merrill et al., 1999; Sharratt and Feng, 2009; Blanco-Canqui et al., 2013), and degraded soil properties (Shaver et al., 2003; Blanco-Canqui et al., 2010) due to less frequent crop growth, lack of residue input, and conventional tillage to manage weeds during fallow. Reduced tillage has been documented to improve soil precipitation storage during the fallow period and may allow for successfully increasing cropping system intensity (Peterson et al., 1996; Saseendran et al., 2009), which could increase profitability and environmental sustainability. Identification of short-season crops to replace or reduce fallow has been a component of many studies (Felter et al., 2006; Lyon et al., 2004, 2007). The impact of replacing fallow on subsequent wheat yield varied significantly but depended primarily on precipitation (Schlegel and Havlin, 1997; Nielsen and Vigil, 2005). Most of these fallow replacement studies evaluated green manure or short-season grain crops. A green manure crop refers to a legume cover crop that is “plowed under and incorporated into the soil,” principally for fixing nitrogen (SSSA, 2017). These studies generally concluded that the short-term benefit of nitrogen fixation from a green manure cover crop or grain production from short-season crops was not enough to compensate for the reduction in soil water and subsequent crop yield (Schlegel and Havlin, 1997; Tanaka et al., 1997; Unger and Vigil, 1998; Nielsen and Vigil, 2005; Allen et al., 2011).

More experimental data are needed that evaluate how replacing fallow with no-till cover crops, annual forages, and grain crops affects soil precipitation storage, subsequent crop yields, and profit. This information is particularly needed for the semiarid Central Great Plains, where moisture is limited and the challenges of successfully increasing cropping intensity are greater. There have been some reports of successful use of cover crops in the Northern Great Plains (Pikul et al., 1997; Burgess et al., 2014), but in the Central and Southern Great Plains, subsequent wheat yields have been reduced more due to greater evapotranspiration (ET) (Nielsen et al., 2015, 2016). However, when cover crops were terminated early enough to allow for a sufficient fallow period, the negative effects on subsequent wheat yields were minimized (Schlegel and Havlin, 1997). Despite the potential negative effects on soil water, producer interest in using cover crops has increased in the semiarid Central Great Plains. However, most producer interest in cover crops is for suppressing herbicide-resistant weeds and forage production (Kansas County Agents, personal communication, 2017). This interest warrants more comprehensive research in terms of benefits to soil and environmental quality, weed management, precipitation storage and use, crop yields, profitability, and long-term effects on soil properties.

Cover crops are not harvested as a cash crop; therefore, the crop and its residue must increase subsequent crop yields, increase nitrogen availability, and/or reduce weeds to justify the expense of planting and growing the cover crop. In contrast, a forage or grain crop grown in place of fallow may have similar beneficial impacts as cover crops but also be harvested as a cash crop. Consequently, at least in the near term, it may be more profitable to grow a forage or grain crop than a cover crop in place of fallow.

Long-term studies are also needed to measure changes over time because soil properties and crop yield response to reduced fallow might improve over time. Although harvesting forage generates economic returns to help offset the costs of growing the crop and yield reductions in subsequent crops (Saseendran et al., 2013; Nielsen et al., 2017), it reduces the amount of residue on the soil surface. Therefore, harvesting a forage crop rather than growing a cover crop may directly undermine the purpose of cover cropping to conserve soil and water resources and enhance system productivity. More information on soil and crop production effects of cover crops in different management scenarios is needed. Our objective was to assess the 5-yr effects of replacing fallow in a no-till W-F rotation with cover crops, annual forages, or grain crops on soil water, wheat yield, grain protein, and profitability in the semiarid Central Great Plains. Our hypothesis was that length of fallow period could be reduced using low-water-use crops terminated several months before wheat planting in no-till W-F while maintaining wheat yields and increasing profitability when harvested for forage.

MATERIAL AND METHODS

Study Description

A long-term experiment of fallow replacement crops (cover crops, annual forages, or grain crops) was located at the Kansas State University Southwest Research and Extension Center near Garden City, KS (37°59’ N, 100°48’ W, 2884 m). The soil series was Ulysses silt loam (fine-silty, mixed, superactive, mesic Aridic Haplustolls) with 1 to 3% slope. The silt loam soil extended to the depth of soil water measurements (1.8 m), with available water capacity being 0.18 m m⁻¹ between field capacity (volumetric water content of 33%) and permanent wilting (volumetric water content of 15%) from the surface to 1.07 m deep and 0.12 m m⁻¹ between field capacity (volumetric water content of 27%) and permanent wilting (volumetric water content of 15%) from 1.07 to 2.6 m deep (Klocke et al., 2013). The semiarid climate had a long-term (1981–2010) average annual precipitation of 489 mm, mean summer growing season daytime high temperature of 29°C (30-yr average, May through August), open-pan evaporation (April through September) of 1810 mm, and a frost-free period of 170 d. The fallow replacement crops consisted of winter and spring crop species that were grown during the fallow period of a no-till W-F cropping system every year from 2007 through 2012 (Table 1). Crops grown in place of fallow were either grown as cover, harvested for forage (annual forage crop), or harvested for grain (spring pea [Pisum sativum L.] and continuous winter wheat). Winter species included hairy vetch (Vicia villosa Roth), lentil (Lens culinaris Medik.), Austrian winter pea, and triticale (×Triticosecale Wittm. ex A. Camus [Secale × Triticum]). Spring species included lentil, field pea, and triticale. Crops were grown in monoculture and in two-species mixtures of each legume plus triticale. Within mixtures, individual species were planted at 50% of monoculture seeding rates (Table 1). Crops grown for grain were continuous winter wheat (W-W) and wheat-field pea (W-GP). Treatments with crops grown in place of fallow were compared with W-F for a total of 27 treatments. The study design was a randomized complete block with four replications and a split-split-plot treatment structure. Crop phase was the main plot, crop species was the split plot, and termination method (cover, forage, or grain) was the split-split plot. Main plot was 138 m wide by 36.6 m long, split plot was 9.2 m wide and 36.6 m long, and split-split plot was 4.6 m wide and 36.6 m long.

Table 1. Fallow replacement crop type, termination method, planting date, termination date, and seeding rate near Garden City, KS

Crop type	Fallow method	Cover crop	Forage crop	Grain crop + other	Planting	Termination	Seeding rate†
		Termination method			Date
							kg ha⁻¹
Other	Fallow	–‡	–	x	–	–	–
Winter	Hairy vetch	x	x	–	1 Oct.	15 May	28
	Hairy vetch–winter triticale	x	x	–	1 Oct.	15 May	67
	Winter lentil	x	x	–	1 Oct.	15 May	39
	Winter lentil–winter triticale	x	x	–	1 Oct.	15 May	78
	Winter pea	x	x	–	1 Oct.	15 May	134
	Winter pea–winter triticale	x	x	–	1 Oct.	15 May	174
	Winter triticale	x	x	–	1 Oct.	15 May	78
	Winter wheat	–	–	x	1 Oct.	1 July	67
Spring	Spring lentil	x	x	–	1 Mar.	1 June	28
	Spring lentil–spring triticale	x	x	–	1 Mar.	1 June	67
	Spring pea	x	x	–	1 Mar.	1 June	134
	Spring pea (grain)	–	–	x	1 Mar.	1 July	78
	Spring pea–spring triticale	x	x	–	1 Mar.	1 June	174
	Spring triticale	x	x	–	1 Mar.	1 June	78

† Mixture components were seeded at 50% of monoculture seeding rates.
‡ –, not applicable; x, applicable.

Each year, winter crops were planted around 1 October, winter cover and forage crops were terminated or harvested around 15 May, and grain crops were harvested around 1 July. Spring crops were planted between the end of February and middle of March, or as early as field conditions would allow planting. Spring cover and forage crops were terminated or harvested around 1 June to minimize negative effects on subsequent wheat yields (Schlegel and Havlin, 1997), and spring field peas grown for grain were harvested around 15 July. Winter and spring cover crop termination coincided with triticale heading (Feekes 10.1; Large 1954), which was selected as a harvest stage to optimize forge yield and quality. Biomass yields for both cover crops and forage crops were determined from a 1-m by 36-m area from within the plot but separate from the area used to measure grain yield and soil water. Biomass was cut to 7.6-cm stubble height using a small plot forage harvester (Carter Manufacturing Company). Total biomass samples were weighed fresh, and subsamples were taken and dried at 50°C for at least 72 h and then weighed dry on consecutive days until a constant weight. After determining biomass yield, a field-scale swather and baler were used to harvest the remaining plot area for the forage treatment. Field peas (grain) and winter wheat were harvested with a small plot combine (Model Delta, Wintersteiger) and stripper header (Model CX, Shelbourne Reynolds) from a 2.4-m by 36-m area at grain maturity, which occurred approximately the first week of July. A stripper header was used to maximize stubble height and residue retention. A grain subsample was collected at harvest and analyzed with a grain analysis computer (Model GAC2100, Dickey John) for moisture and test weight. Grain yield was adjusted to 13.5% moisture content. Grain samples were analyzed for nitrogen content by an automated sulfuric peroxide digestion (Linder and Harley, 1942; Thomas et al., 1967), and multiplied by 5.7 to convert nitrogen content to grain protein (Jones, 1941; AOAC, 1984). Treatment effects on soil physical properties were previously reported (Blanco-Canqui et al., 2013).

Measurement of Available Soil Water

Volumetric soil water content was measured at planting and at harvest of winter wheat and fallow replacement crops (cover crops, forage crops, and grain crops) by gravimetric method using a 2.22-cm-diam. soil core probe (Model GSRPS, Giddings Machine Company) in six 0.3-m increments to a 1.8-m soil depth. In addition, volumetric soil water content was measured in the 0.00- to 7.62-cm soil depth at wheat planting to quantify soil water in the seed placement zone by compositing four 0.25-cm-diam. cores per plot by gravimetric method using a hand probe (Lord Soil Probe). Soil water samples were processed and converted to plant available water (PAW) as described by Black (1965).

Seasonal crop water use or ET was estimated as a residual of the soil water balance as expressed in Eq. [1]:

$urn:x-wiley:2578-2703:equation:csc2cropsci2017050324-math-0001$ (1)

where I is irrigation (mm), which is zero since no irrigation was applied; P is precipitation (mm); R is runon or runoff, assumed negligible (mm); SW1 is soil water at wheat planting; SW2 is soil water measured at wheat harvest maturity (mm); and D is deep drainage, estimated according to Stone et al. (2011). Water balance was also evaluated for crops used in fallow replacement. Water use efficiency (WUE) was estimated as the ratio of economic yield to seasonal ET or crop water.

Partial Budget Analysis

A partial budget analysis of all cropping systems (fallow, cover crop, forage crop, and grain crop) was compared across years (2008–2012). Production costs were derived from reported custom rates in Kansas (Ibendahl, 2016), wheat production budgets (Ibendahl et al., 2016), and local fertilizer, seed, grain, hay, and herbicide prices. Grain and forage prices were the 5-yr average reported local market prices (USDA-AMS, 2017; USDA-NASS, 2017). No premium or discount was attributed for grain protein because, unlike in some other regions, grain price in Kansas is rarely affected by protein content.

Statistical Analysis

The data analysis for cover and forage crop biomass production, wheat (protein, test weight, yield, and WUE), PAW, and net return was performed using PROC MIXED, residual maximum likelihood method (SAS Institute, 2012). Replication and replication × treatment interactions were considered random effects, and all other effects, including year, were considered fixed in the model. Wheat yield from 2008 was excluded from the analysis due to hail damage that year. Unless otherwise indicated, data were analyzed separately by year due to significant fixed effect interactions with year. Tests of fixed effects used the Satterthwaite approximation for the denominator degrees of freedom. Cover and forage crop were compared using a single degree of freedom contrast. Treatment effects were considered significant at P ≤ 0.05, and least squares means were separated with significance level of P ≤ 0.05 in PROC MIXED (PDIFF option). Regression analysis was used to characterize the relationship between biomass production, soil water, and grain yield with PROC REG (SAS Institute, 2012).

RESULTS AND DISCUSSION

Climatic Conditions

During the study (2007–2012), average annual precipitation was measured at 429 mm from a weather station 400 m from the study site. Precipitation received during the winter wheat growing season during the 4 yr reported of this study ranged from 172 mm in 2011 to 412 mm in 2009. The precipitation received during the 15-mo fallow period of W-F ranged from 366 mm in 2010 to 702 mm in 2008, and precipitation during the 90-d fallow period of W-W ranged from 85 mm in 2011 to 177 mm in 2009 (Table 2). The 30-yr average precipitation for the research location was 318 mm during the wheat growing season, 660 mm during the 15-mo fallow period, and 171 mm during the 3-mo fallow period. These conditions provided a broad range of water availability for quantifying cover, forage, and grain production across the range of fallow replacement treatments.

Table 2. Winter wheat growing season and fallow precipitation near Garden City, KS

Growing season	Growing season (Oct.–June)	Preceding 15 mo fallow (July–Dec. plus Jan.–Sept.)	Preceding 3 mo fallow (July–Sept.)
	Precipitation
	mm
2008–2009	412	702	113
2009–2010	358	644	177
2010–2011	172	366	109
2011–2012	216	423	85
30 yr avg.†	318	660	171

† 30-yr average precipitation for the growing season and fallow periods.

Annual Forage and Cover Crop Biomass

Biomass production was similar for treatments managed as a cover crop (left standing) or annual forage; therefore, biomass accumulation was averaged across the two management treatments. Biomass production was least (780 kg ha⁻¹ treatment average) in 2011 (Table 3), when precipitation during the cover crop growing period was 46% less than the 30-yr average, and greatest in 2008 and 2010 (2690 kg ha⁻¹ treatment average), when precipitation was at or above the 30-yr average (Table 2). Winter triticale and winter triticale–legume mixtures produced the most biomass, averaging 4100 kg ha⁻¹, or 140% more than the second greatest biomass treatment of spring triticale and spring triticale–legume mixtures (1700 kg ha⁻¹ treatment average) (Table 3). Although individual species biomass accumulation within a mixture was not quantified, triticale comprised the majority of the biomass in mixture treatments. Legumes produced the least biomass, averaging 714 kg ha⁻¹. The exception was in 2011, when spring triticale produced more than winter triticale due to dry conditions in the fall and winter of 2010, followed by 120 mm of precipitation from April through June 2011. This greater spring precipitation benefited the growth of spring-planted cover or forage crops in 2011. Hairy vetch and winter pea often had winterkill and stand loss when grown in monoculture, resulting in no hairy vetch biomass in 2009 or 2011 and no winter pea biomass in 2011. However, no winterkill was observed when hairy vetch or winter pea were grown in a mixture with winter triticale, even resulting in increased biomass production of the winter triticale–hairy vetch mix (1130 kg ha⁻¹) compared with winter triticale monoculture (950 kg ha⁻¹) in 2011. Winter lentil, unlike winter pea and hairy vetch, consistently had good winter survival and produced biomass similar to that produced by spring lentil at 500 kg ha⁻¹. Previous studies found Austrian winter pea survival to be potentially reduced by tall no-till residue, yet winter lentil survival was potentially increased by tall residue (Huggins and Pan, 1991; Chen et al., 2006). The no-till cropping system used in this study might have negatively affected winter pea and hairy vetch survival and improved winter lentil survival. The mechanism is uncertain, but the microenvironment in triticale mixture plantings increased winter legume survival.

Table 3. Cover crop and forage crop biomass by treatment and year near Garden City, KS

Source	Fallow method	2008	2009	2010	2011	2008–2011 avg.
		kg ha⁻¹
Crop type
Winter	Hairy vetch	1052e†	0e	590de	0d	438d
	Hairy vetch–winter triticale	4380a	5042a	5333a	1132abc	4161a
	Winter lentil	512e	300e	529e	28d	301d
	Winter lentil–winter triticale	4026ab	4147b	5523a	976bc	3848a
	Winter pea	982e	597de	751de	0d	622d
	Winter pea–winter triticale	4144a	5845a	5139a	873c	4209a
	Winter triticale	3638abc	5292a	5794a	947bc	4116a
Spring	Spring lentil	874e	410e	500e	195d	515d
	Spring lentil–spring triticale	2606cd	1417cd	1891bc	1222abc	1822bc
	Spring pea	1795de	1643c	1565cd	1415ab	1601c
	Spring pea–spring triticale	2812cd	2032c	2727b	1009bc	2245b
	Spring triticale	3024bc	1659c	1984bc	1575a	2093b
P-value		****	****	****	****	****
LSD_0.05		1111	890	1002	472	469
Year	2008					2688a
	2009					2372b
	2010					2692a
	2011					778c
P-value						****
LSD_0.05						271

**** Significant at the 0.0001 probability level.
† Within columns, means followed by same letter are not significantly different according to LSD_0.05.

Plant Available Water

Plant available water at the depth of seed placement (0.0 to 7.6 cm) was critical for wheat germination and seedling establishment, and PAW in the soil profile (0.0 to 1.8 m) was critical for grain yield (Norwood, 1994; Nielsen et al., 2002; Holman et al., 2011). Plant available water at wheat planting was affected greatly by year of the study (Table 4). Plant available water in the top 7.6 cm of the soil profile ranged from 0 mm in 2011 to 7 mm in 2010, and in the 0.0- to 1.8-m soil profile, 72 mm in 2012 to 193 mm in 2010 (Table 4). Previous research in the region determined that seed zone moisture and precipitation near planting accounted for more variation in wheat yield than precipitation or temperature received at any other time during the growing season (Holman et al., 2011). Stone and Schlegel (2006) determined that PAW and growing season precipitation each accounted for 30% of the variability in wheat yield and together explained ∼70% of the variability in wheat yield. The 2009 and 2010 growing seasons had the most precipitation and initial PAW in the top 7.6-cm soil profile, and the 2011 and 2012 seasons had the least precipitation and PAW in the top 7.6-cm soil profile (Table 3). These differences in PAW and growing season precipitation resulted in a range of outcomes of growing crops (cover crop, forage crop, or grain) in place of fallow on subsequent wheat yields.

Table 4. Plant available water (PAW) at wheat planting averaged across treatments by year, and averaged across years and treatments terminated as either a cover crop or forage crop near Garden City, KS

Cover crop method	PAW (0.0–7.6 cm)	PAW (0.0–1.8 m)
	Mm
Year
2008–2009	0.9b†	111c
2009–2010	7.0a	193a
2010–2011	0.0d	148b
2011–2012	0.1c	72d
P-value	****	****
LSD_0.05	0.7	12
Termination method
Cover crop	2.4a	146a
Forage crop	1.5b	126b
P-value	***	****
LSD_0.05	0.7	11

*** Significant at the 0.001 probability level.
**** Significant at the 0.0001 probability level.
† Within columns, means followed by same letter are not significantly different according to LSD_0.05.

Managing biomass crops as a cover crop increased PAW in the top 7.6-cm soil profile by 0.9 mm and in the 0.0- to 1.8-m soil profile by 20 mm compared with managing as a forage crop (Table 4). The increase in PAW with a cover crop was attributed to reduction in evaporation rates that resulted in an increase in precipitation storage efficiency between termination and wheat planting because PAW at cover crop and forage termination were similar. There was no interaction between management and crop species on the amount of PAW.

Plant available water was affected by crop species and year, but crop species × year interaction effect was not significant. Winter triticale, winter triticale–legume mixtures, and spring triticale had more PAW in the top 7.6 cm of soil profile than fallow and other crops (Table 5). The greater PAW in the upper soil surface of these treatments was attributed to a drier soil profile and more aboveground biomass increasing precipitation storage efficiency (Peterson and Westfall, 2004; Nielsen et al., 2005). Even though forage biomass was harvested, the remaining stubble often had more surface residue cover than fallow. The exception was in 2011, when the process of planting a fallow crop destroyed previous wheat residue and grew little to no biomass due to very dry conditions (Table 3). Previous studies have shown the importance of maintaining or increasing surface residue to reduce soil water evaporation (Norwood, 1994; Peterson and Westfall, 2004; Nielsen et al., 2005), and the present study supports those findings. Plant available water at wheat harvest was low and similar across treatments (data not shown). However, at wheat planting, fallow had more PAW at all depths except 0.0 to 30.5 cm, a result of less soil water used and more soil water stored compared with growing a cover, forage, or grain crop during the fallow period (data not shown, P ≤ 0.0001). Grain crops grown during fallow, W-GP, and W-W tended to have the least PAW at all depths. Continuous wheat had less PAW than W-GP at soil depths below 91.5 cm, which was likely due to less pea biomass and grain yield than winter wheat. Winter triticale and winter triticale–legume mixtures had less PAW than spring triticale and spring triticale–legume mixtures at soil depths below 122 cm. This occurred because of a shorter growing season, less biomass production, and less water used by spring triticale compared with winter triticale. Winter triticale water use was similar to that of winter wheat, although winter triticale left more PAW at soil depths below 91.5 cm. Legumes tended to leave more PAW than triticale or grain crops but less PAW than fallow at all depths. The greater moisture content at deeper depths in fallow was due to more soil water storage throughout the fallow period, which benefited wheat growth later in the growing season.

Table 5. Plant available water (PAW) at wheat planting in the 0.0- to 7.6-cm and 0.0- to 1.8-m soil depths near Garden City, KS

Crop type	Fallow method	PAW (0.0–7.6 cm)	PAW (0.0–1.8 m)
		Mm
Other	Fallow	1.5cde†	201a
Winter	Hairy vetch	1.6cde	158b
	Hairy vetch–winter triticale	3.4a	131def
	Winter lentil	1.0ef	154bc
	Winter lentil–winter triticale	2.4abc	114efg
	Winter pea	1.7bcde	137bcd
	Winter pea–winter triticale	3.0ab	126defg
	Winter triticale	2.3abcd	109fg
	Winter wheat (grain)	−0.3fg	83h
Spring	Spring lentil	1.1de	144bcd
	Spring lentil–spring triticale	1.6cde	131cdef
	Spring pea	1.4cde	157b
	Spring pea (grain)	−0.5g	104gh
	Spring pea–spring triticale	1.6cde	133cde
	Spring triticale	1.8bcde	139bcd
P-value		***	****
LSD_0.05		1.3	23

*** Significant at the 0.001 probability level.
**** Significant at the 0.0001 probability level.
† Within columns, means followed by same letter are not significantly different according to LSD_0.05.

Cover crops or forages that grew the most biomass did so by using the most water, resulting in the least amount of PAW at wheat planting in the 0.0- to 1.8-m soil profile (Fig. 1). For every 125 kg ha⁻¹ biomass produced, PAW was reduced 1 mm (Fig. 1). Although precipitation storage after cover crop termination was increased with surface residue, there was insufficient soil water storage to compensate for the water used to grow the crop in this study. Fallow had the most PAW, averaging 200 mm, followed by legumes with 150 mm, spring triticale and spring triticale–legume mixtures with 135 mm, winter triticale and winter triticale–legume mixtures with 120 m, W-GP with 100 mm, and lastly W-W with 80 mm (Table 5). Crops grown in place of fallow for grain production in W-GP and W-W had a longer growing season and shorter fallow period (90 d) than crops grown for biomass only (120–135 d), resulting in the lowest PAW at wheat planting (Table 5). Lyon et al. (2007) suggested that forage crops might be the best option for a flexible-fallow planting due to the least amount of soil water use. Spring grain crops in the Southern Great Plains often experience heat stress during flowering, resulting in reduced yields, and in this study, pea grain yield averaged only 850 kg ha⁻¹. Better pea grain yields might have resulted in less PAW and greater net return than observed in this study.

Details are in the caption following the image — **Fig 1**
Open in figure viewer

Impact of cover crop and forage crop biomass production on subsequent plant available water at wheat planting in the 0.0- to 1.8-m soil depth from 2008 through 2012.

Winter Wheat Production

Wheat yield was affected by growing season (P ≤ 0.0001) and crop species (P ≤ 0.0001) grown in place of fallow but was not affected by fallow crop management (i.e., cover crop vs. forage crop). Wheat yield averaged 5500 kg ha⁻¹ in 2009, 4450 kg ha⁻¹ in 2010, 1240 kg ha⁻¹ in 2011, and 633 kg ha⁻¹ in 2012. This corresponded well with growing season precipitation and measured PAW at wheat planting (Tables 2 and 4). Due to differences in PAW and growing season precipitation, there was a significant year × crop species interaction effect on wheat yield.

With favorable growing conditions such as in 2009 and 2010, wheat yield potential after fallow was 4700 kg ha⁻¹ or greater. Under these conditions cover crops grown in place of fallow had little effect on wheat yield compared with fallow (Table 6). Wheat grown after legumes that used little water during the fallow period (Tables 5 and 6) yielded similar to that grown after fallow. Wheat grown after winter lentil and spring pea yielded 8% more than after fallow in 2009, which was the only time during the course of this study that fallow did not result in the greatest wheat yield. High-water-use cover crops and forage crops reduced wheat yield an average of 6% after winter triticale and winter triticale–legume mixtures and 4% after spring triticale and spring triticale–legume mixtures between 2009 and 2010 (Table 6). Peas harvested for grain reduced wheat yield 8% and W-W reduced wheat yield 34% during this same time period. Pea grain yields were low and often failed to produce grain. Had pea grain yields been greater, the negative impact on wheat yield likely would have been greater and more similar to W-W.

Table 6. Winter wheat yield by crop species and year near Garden City, KS

Source	Fallow method	2009	2010	2011	2012	2009–2012 avg.
		kg ha⁻¹
Crop type
Other	Fallow	5590cdef†	4699ab	1571abc	2157a	3739a
Winter	Hairy vetch	5789abcd	4818a	1122defgh	748bcd	3466bcd
	Hairy vetch–winter triticale	5437defg	4455abc	1158defgh	401de	3160efg
	Winter lentil	6121a	4688ab	1491abcd	824bc	3584ab
	Winter lentil–winer triticale	5208fg	4548abc	1240cdefg	677bcde	3124fg
	Winter pea	5831abcd	4688ab	1098efgh	414cde	3329cdef
	Winter pea–winter triticale	5242fg	4382bc	898ghi	330e	2962gh
	Winter triticale	5209fg	4268c	1269cdefg	583bcde	3105fg
	Winter wheat (grain)	3848h	2924d	585i	818bc	2236i
Spring	Spring lentil	5951abc	4574abc	1707a	963b	3543abc
	Spring lentil–spring triticale	5294efg	4412bc	1470abcde	569bcde	3275def
	Spring pea	6054ab	4518abc	1057fgh	541cde	3363bcde
	Spring pea (grain)	5082g	4387bc	823hi	279e	2759h
	Spring pea–spring triticale	5588cdef	4285c	1319bcdef	519cde	3215ef
	Spring triticale	5683bcde	4524abc	1686ab	526cde	3390bcde
P-value		****	****	****	****	****
LSD_0.05		403	403	379	414	239
Year	2009					5498a
	2010					4454b
	2011					1242c
	2012					633d
P-value						****
LSD_0.05						100

**** Significant at the 0.0001 probability level.
† Within columns, means followed by same letter are not significantly different according to LSD_0.05.

In the drier growing seasons of 2011 and 2012, growing a crop in place of fallow had a greater negative effect on wheat yield (Table 6). The decrease in wheat yields was more pronounced in 2012, when the fallow period did not receive sufficient precipitation to supply PAW at wheat planting (Tables 4 and 6). During 2011 and 2012, winter wheat yields were reduced 47% by legumes, 46% by spring triticale and triticale–legume mixtures, 56% by winter triticale and triticale–legume mixtures, 70% by W-GP, and 62% by W-W. In 2012, W-W yielded more than W-GP or winter triticale–legume mixtures, which was likely due to more resilient wheat residue reducing soil water evaporation and increasing wheat WUE (Table 7). During these dry growing seasons, wheat yield was correlated to PAW (Fig. 2), but in the wet growing seasons, PAW was not correlated to wheat yield. Lyon et al. (2007) reported a similar winter wheat yield relationship to PAW. As described previously, PAW was negatively correlated with cover and forage biomass production. Averaged across years, wheat yield was greatest after fallow, with yields after winter or spring lentil being similar to fallow (Table 6). Wheat yields were reduced 12% after a cover or forage crop, 21% after W-GP, and 40% after W-W. Across years, the relationship of winter wheat yield was 12.95 kg ha⁻¹ mm⁻¹ of ET (Fig. 3), similar to other studies in the region (Nielsen et al., 2002; Stone and Schlegel, 2006).

Table 7. Winter wheat water use efficiency (WUE) by crop species and year near Garden City, KS

Source	Fallow method	2009	2010	2011	2012	2009–2012 avg.
		WUE
		kg ha mm⁻¹
Crop type
Other	Fallow	1.6	1.3b†	0.7ab	1.2a	1.3ab
Winter	Hairy vetch	1.9	1.4ab	0.6bcd	0.5bcd	1.2abcde
	Hairy vetch–winter triticale	1.8	1.4ab	0.7abc	0.3de	1.1efg
	Winter lentil	1.9	1.5a	0.7ab	0.5bcd	1.2abc
	Winter lentil–winer triticale	1.8	1.4ab	0.7ab	0.4bcde	1.1cdef
	Winter pea	2.0	1.4ab	0.6bcd	0.3de	1.2bcde
	Winter pea–winter triticale	1.8	1.4ab	0.5cde	0.2e	1.1fg
	Winter triticale	1.8	1.3b	0.7ab	0.4bcde	1.1cdef
	Winter wheat (grain)	1.8	1.1c	0.4e	0.7b	1.0g
Spring	Spring lentil	2.0	1.5a	0.9a	0.6bc	1.3a
	Spring lentil–spring triticale	1.8	1.4ab	0.7abc	0.4bcde	1.2cdef
	Spring pea	1.8	1.4ab	0.6bcd	0.4cde	1.1efg
	Spring pea (grain)	2.0	1.4ab	0.5de	0.2e	1.1fg
	Spring pea–spring triticale	1.8	1.3b	0.7abc	0.4cde	1.1def
	Spring triticale	2.0	1.4ab	0.9a	0.4bcde	1.2abcd
P-value		NS‡	**	**	****	****
LSD_0.05		–	0.1	0.2	0.3	0.1
Year	2009					1.8a
	2010					1.4b
	2011					0.7c
	2012					0.4d
P-value						****
LSD_0.05						0.1

** Significant at the 0.01 probability level.
**** Significant at the 0.0001 probability level.
† Within columns, means followed by same letter are not significantly different according to LSD_0.05.
‡ NS, nonsignificant.

Water use efficiency varied by growing season and soil water conditions. There were no differences among treatments in 2009 (Table 7). In 2010, WUE was least in W-W and greatest in winter or spring lentil. In 2011, WUE was least in W-W, W-GP, and winter triticale–winter pea and greatest in winter lentil, spring lentil, and spring triticale. In 2012, WUE was lower due to drought conditions and was greatest in W-F. Water use efficiency after cover crops was similar to fallow when growing season moisture was plentiful, but in dry years, WUE after cover crops was less than fallow. Treatments with low PAW such as W-W or W-GP tended to have the lowest WUE due to insufficient PAW at the beginning of the growing season. Averaged across years, WUE was greatest in fallow and in winter or spring lentil treatments—those treatments with the highest PAW. There were no differences in WUE among forage or cover crop treatments. There was no improvement in wheat yield response to growing a cover or forage crop in place of fallow, and yield was dependent on PAW and growing season precipitation. These findings are similar to other cover crop research in the semiarid Central Great Plains region (Nielsen et al., 2015, 2016).

Winter wheat grain test weight and protein content were an indication of crop stress during grain fill and correlated closely with grain yield. Test weight was lowest in 2011 and 2012 (74 kg hL⁻¹), when growing season precipitation was least, and was greater in 2009 (76.5 kg hL⁻¹) and 2010 (74.5 kg hL⁻¹), when growing season conditions were more favorable. Test weight tended to be greatest for wheat after fallow and legumes and least in W-W (data not shown, P ≤ 0.0001). Neither test weight nor protein content was affected by management (cover crop vs. forage crop). Protein content was greater in the drier growing seasons such that 2012 (186 g kg⁻¹) > 2011 (175 g kg⁻¹) > 2009 (155 g kg⁻¹) > 2010 (139 g kg⁻¹) (P ≤ 0.0001). Dry growing conditions and stress during grain fill that were observed in this study and other studies have reported increased protein and lower test weight due to smaller, less dense endosperm (Lee et al., 2013). Wheat grain protein content <121 g kg⁻¹ (at 13 g kg⁻¹ moisture content) indicated potential yield-limiting soil nitrogen levels (Engel et al., 2005; Jones and Olson-Rutz, 2012). Therefore, wheat yields in this study were likely not limited by soil nitrogen levels and most likely not influenced by any potential nitrogen provided by legumes grown in the rotation, although it did appear that legumes grown in the rotation tended to increase grain protein content (160 to 166 g kg⁻¹) compared with fallow (155 g kg⁻¹) (data not shown, P ≤ 0.0001). Continuous winter wheat had the greatest protein content (168 g kg⁻¹), which was likely due to W-W fertilized similar to W-F and excess nitrogen fertilizer being available for the amount of grain produced.

Profitability

Using a partial budget analysis for the fallow phase of the cropping system between 2008 and 2012, gross and net returns for all treatments were determined (Table 8). The partial budget accounted for only variable costs that were affected by fallow management including herbicide application, seeding and harvesting the fallow-replacement crops, and effects on subsequent winter wheat yields. This partial budget excluded land rent, crop insurance, interest, and variable or fixed costs not affected by fallow management (Ibendahl et al., 2016). During the course of this study, one less herbicide application was required when a crop was grown in place of fallow due to weed suppression by the crop (Petrosino et al., 2015). The impact of growing a crop in place of fallow on subsequent wheat yields was influenced by precipitation amount, which in turn also affected annual economic returns. However, it was assumed that a producer would adopt a specific fallow practice for all years, since future precipitation levels were unknown. During the course of this study, there were two growing seasons with above- and two growing seasons with below-normal precipitation, which allowed for a reasonable estimate of long-term economic returns. A nitrogen credit was not included in the economic analysis, since neither soils nor grain were affected by legumes in this study. Producers spend money on herbicides to manage weeds in fallow. Therefore, alternative fallow management systems could have a negative cash flow yet be more profitable if they lost less money than the control fallow treatment. Lyon et al. (2004) found that wheat yields might be reduced by growing a crop in place of fallow, yet the cropping system might be more profitable if the returns during the fallow period were increased.

Table 8. Partial enterprise analysis of the fallow phase in a winter wheat-fallow crop rotation growing a cover crop, forage crop, spring grain peas, or continuous winter wheat from 2009 through 2012 near Garden City, KS

Accounting ledger		Fallow	Hairy vetch	Hairy vetch– triticale	Lentil	Lentil–triticale	Pea	Pea–triticale	Triticale	Wheat	Lentil	Lentil–triticale	Pea	Pea (grain)	Pea–triticale	Triticale
		Fallow management practice
		Other	Winter								Spring
Expenses	Seeding, US$ ha⁻¹	–	49	49	49	49	49	49	49	49	49	49	49	49	49	49
	Seed cost, US$ ha⁻¹	–	139	87	30	33	62	49	36	21	26	35	68	68	56	43
	Total seeding cost, US$ ha⁻¹	–	188	136	79	82	111	98	85	70	75	83	117	117	105	92
	Swathing, US$ ha⁻¹	–	43	43	43	43	43	43	43	–	43	43	43	–	43	43
	Baling and stacking, US$ ha⁻¹	–	0.03	0.03	0.03	0.03	0.03	0.03	0.03	–	0.03	0.03	0.03	–	0.03	0.03
	Total hay cost, US$ ha⁻¹	–	59	189	54	178	65	191	187	–	61	107	99	–	122	117
	Grain harvesting, US$ ha⁻¹	–	–	–	–	–	–	–	–	56	–	–	–	56	–	–
	Grain hauling, US$ ha⁻¹	–	–	–	–	–	–	–	–	0	–	–	–	0	–	–
	Total grain harvesting cost, US$ ha⁻¹	–	–	–	–	–	–	–	–	73	–	–	–	60	–	–
	Fallow herbicide application, US$ ha⁻¹	16	16	16	16	16	16	16	16	16	16	16	16	16	16	16
	Fallow herbicide per application, US$ ha⁻¹	18	18	18	18	18	18	18	18	18	18	18	18	18	18	18
	Fallow applications per period	4	3	3	3	3	3	3	3	3	3	3	3	3	3	3
	Total fallow herbicide cost, US$ ha⁻¹	135	101	101	101	101	101	101	101	101	101	101	101	101	101	101
	In-crop herbicide application, US$ ha⁻¹	–	–	–	–	–	–	–	–	16	–	–	–	16	–	–
	In-crop herbicide per application, US$ ha⁻¹	–	–	–	–	–	–	–	–	15	–	–	–	15	–	–
	In-crop applications per period	–	–	–	–	–	–	–	–	1	–	–	–	1	–	–
	Total in-crop herbicide cost, US$ ha⁻¹	–	–	–	–	–	–	–	–	30	–	–	–	30	–	–
Total expenses (cover), US$ ha⁻¹		–	289	238	180	183	212	199	186	–	176	185	219	–	206	194
Total expenses (hay), US$ ha⁻¹		–	348	427	234	361	277	390	374	–	238	292	318	–	328	310
Total expenses (other), US$ ha⁻¹		135	–	–	–	–	–	–	–	275	–	–	–	309	–	–
Income	Hay production, kg ha⁻¹	–	577	5488	398	5074	820	5551	5428	–	679	2402	2112	–	2960	2760
	Grain yield, kg ha⁻¹	–	–	–	–	–	–	–	–	2220	–	–	–	538	–	–
	Price, US$ kg⁻¹	–	0.09	0.09	0.09	0.09	0.09	0.09	0.09	0.13	0.09	0.09	0.09	0.26	0.09	0.09
	Production proceeds, US$ ha⁻¹	–	51	484	35	447	72	489	479	285	60	212	186	138	261	243
	Effect on subsequent wheat crop, kg ha⁻¹	0	−386	−643	−224	−587	−498	−793	−673	−1464	−206	−564	−463	−863	−578	−401
	Effect on subsequent wheat crop, US$ ha⁻¹	0	−50	−83	−29	−76	−64	−102	−87	−188	−26	−72	−59	−111	−74	−52
Net return (cover), US$ ha⁻¹		–	−339	−320	−209	−259	−276	−301	−273	–	−203	−257	−278	–	−280	−245
Net return (hay), US$ ha⁻¹		–	−347	−26	−228	11	−269	−2	18	–	−204	−153	−191	–	−141	−118
Net return (other), US$ ha⁻¹		−135	–	–	–	–	–	–	–	−178	–	–	–	−282	–	–
LSD_0.05 (net return)†		52

† Net return at LSD 0.05 for comparison across fallow management practice (cover, hay, or other) and crop species.

In the present study, growing winter triticale or winter triticale–legume forage during the fallow period was the most profitable treatment across all years, with net returns for the period ranging from −US$25.53 to $18.42 ha⁻¹ (Table 8). Fallow had the next highest return (−$135.21 ha⁻¹), which was similar to returns from spring triticale or spring triticale–legume forage ranging from −$152.49 to −118.15 ha⁻¹. In 2012, spring triticale and spring triticale–legume forage were less profitable than fallow due to insufficient forage production to offset the greater reduction in wheat yield during a dry year. However, from 2009 to 2011 during favorable growing conditions, spring triticale forage was more profitable than fallow (data not shown). Growing a forage crop in wet years and using fallow in dry years (i.e., “flex-fallow”) would provide the highest returns. Unfortunately, precipitation outlook models provide a very poor estimate of actual rainfall (CPC, 2017). Legume-only forage treatments were less profitable than fallow or triticale forage due to low legume biomass production, with returns ranging from−$346.40 (vetch) to −$190.99 (spring pea) ha⁻¹. Triticale–legume forage treatments tended to have less profit than triticale-only treatments due to little or no forage yield increase with the inclusion of a legume in the mixture and the higher cost of legume seed. The particularly high seed cost of hairy vetch resulted in it being the least profitable treatment. Continuous wheat and W-GP were less profitable than W-F due to lower wheat yields and low pea grain yield. Continuous winter wheat was more profitable than W-F only in 2009 under wet growing conditions. Pea grown for grain in the Southern Great Plains experiences heat stress during flowering and grain fill, which can greatly reduce yield. Including pea for grain production in the Northern Great Plains, where pea is more commonly grown, likely would have a greater profit potential than observed in this study. Cover crop treatments were less profitable than fallow or forage treatments and on average were 100% more expensive to implement than fallow. A cover crop of hairy vetch was 150% less profitable than fallow. Of the cover crop treatments, winter and spring lentils were the most profitable due to the least amount of water used and least negative impact on wheat yield, yet lentil cover crops reduced profit 52% more than fallow. Averaged across years, spring triticale forage increased net returns 26% and winter triticale forage increased net returns 240% compared with fallow.

CONCLUSIONS

Growing a crop in place of fallow reduced PAW at wheat planting. Cover crops had slightly more PAW than forage crops, but not enough of a difference to affect wheat yield. Cover crops and forage crops that grew more biomass used more soil water, which resulted in lower wheat yields. A grain crop grown in place of fallow had a shorter fallow period than a cover or forage crop, resulting in even greater wheat yield reductions. Wheat yields were reduced 5.5 kg ha⁻¹ for every millimeter less PAW at wheat planting. In wet years, growing season precipitation was sufficient to minimize yield reductions from growing a cover or forage crop during the fallow period. However, in dry years, growing a crop during the fallow period reduced wheat yields by as much as 70%. Results from this study suggest that a cover or forage crop could be grown in place of fallow in a W-F rotation with minimal impacts on wheat yield, if there is sufficient growing season precipitation to support wheat yield potential of 3500 kg ha⁻¹ or greater.

Improving soil properties and productivity is a lofty goal; however, if it is not profitable, it will be unachievable. Cover crops greatly reduced net returns compared with fallow; however, growing a forage crop in place of fallow increased net return when the forage crop produced sufficient forage biomass accumulation and did not have high seed cost. Producers wanting to make a profit should either use fallow or grow a forage crop like winter or spring triticale and not grow species mixtures with high seed cost or low forage biomass accumulation.

Other findings from this study previously reported growing a cover crop or forage in place of fallow improved soil organic carbon and aggregate stability after 5 yr of growing cover crops in place of fallow, but the effects on soil properties lasted <9 mo after the cover crop was terminated (Blanco-Canqui et al., 2013). The long-term effects (>5 yr) of growing a crop in place of fallow on soils and grain yield should be evaluated, as well as comparing forage crops harvested for hay or grazed. In order for producers to remain profitable and stay in operation, however, net returns of alternative production systems must be equal to or greater than those of fallow. These results suggest that the best way for producers to potentially improve soil properties over time and be profitable is to grow a forage crop and not a cover crop in the semiarid Central Great Plains region.

Conflict of Interest

The authors declare that there is no conflict of interest.

Acknowledgments

This material was based on work supported by the Cooperative State Research, Education, and Extension Service, USDA, under Awards no. 2007-34103-18104, 2007-41530-03801, and 58-3090-5-007 and the Kansas Agricultural Experiment Station. Contribution no. 17-353-J from the Kansas Agricultural Experiment Station.

References

Citing Literature

Volume58, Issue2

March–April 2018

Pages 932-944

Journal list menu

Tools

Follow journal

Can Cover or Forage Crops Replace Fallow in the Semiarid Central Great Plains?