1 Introduction

Alfalfa is one of the world’s most versatile crops, grown in environments ranging from burning hot deserts to cool high mountain valleys (Li and Su 2017). Many farmers and cultures value its high productivity, flexible wide adaptation to different soil types and climatic zones, and life-sustaining nutritional characteristics (Das et al. 2021). This adaptability occurs as a result of some strategies that alfalfa develop to sustain its growth in a wide range of environments, such as modification of its leaf area ratio or increasing shoot to root ratio to allow the roots to capture more water during drought (Anower et al. 2017). Consequently, this depletes underground water and water storage reservoirs rapidly, and continuous alfalfa cultivation can culminate in extreme soil desiccation (Das et al. 2021).

Drought is one of the most crucial factors limiting the growth and production of most economic crops worldwide. Therefore, proper soil management practices can be adopted for growing alfalfa in sunken beds for the utilization of seepage water from raised beds (Mak-Mensah et al. 2021a). A simple land configuration through ridge and furrow system is a useful technique for proper land and water management to increase crop water productivity (Amarasingha et al. 2017). Modification in field topography through the construction of alternate ridges and furrows improves the physical environment, particularly the aeration status of the soil, and creates conducive conditions for crop growth (Das et al. 2021). Water management practices like “ridge-furrow rainwater harvesting” (RFRH) tackle water scarcity in a better way in rain-fed alfalfa cultivation and has the potential to contribute to more sustainable water use efficiency (Mak-Mensah et al. 2021b).

One of the most effective techniques in soil and water loss is mulching (Liu et al. 2018a). Ridge mulching increases soil temperature and precipitation infiltration and decreases crop water consumption (Mo et al. 2016). Mulching typically with plastic film, crop straw, animal manure, gravel-sands, and rocks are applied to fields before, during, or shortly after sowing (Yang et al. 2020). However, widespread materials applied in RFRH for mulching are straw and plastic film (Li et al. 2017a). Plastic films decrease soil water evaporation and increase crop transpiration, thus promoting crop yield (Mak-Mensah et al. 2021a). Despite these benefits, residual plastic film deposits adversely affect soil structure and transport of nutrients and water (He et al. 2018a).

Prior investigations of ridge-furrow rainwater harvesting systems in northwestern China have evaluated effects of agronomic practices on yield and water use efficiency in alfalfa production (Fan et al. 2019; He et al. 2018b; Jia et al. 2018; Li et al. 2017b; Zhang et al. 2021). However, to date, no evaluation has been conducted in the region to assess effects of ridge-furrow rainwater harvesting system with mulching on fodder yield, quality, soil desiccation, and their economic benefits. Therefore, this study was undertaken to evaluate the effects of ridge-furrow rainwater harvesting system with mulching on alfalfa yield, fodder quality, soil desiccation, and their economic benefits in northwestern China over 5 years. We anticipated that costs from mulching material would be compensated by improved alfalfa yield and fodder quality under ridge-furrow rainwater harvesting system and mulching compared to flat planting.

2 Materials and Methods

2.1 Study Area

Field experiments were conducted from 2012 to 2016 at Dingxi Arid Meteorology and Ecological Environment Experimental Station, Institute of Arid Meteorology of China Meteorological Administration. The station is located 3 km southeast of Dingxi City in Gansu Province, Northwest of China (35°34′59″ N 104°37′00″E, with an elevation of 1971 m a. s. l.). Dingxi is a semi-arid area with small farming villages sited in narrow valleys surrounded by mountains with a mean annual air temperature of 7.2 °C (Fig. 2a). During study seasons, the lowest and the highest annual mean temperatures were − 13.0 °C and 25.9 °C, respectively. Approximately 80% of rainfall, which contributed to annual mean rainfall of 408 mm, mainly occurred from May to September. Annual evapotranspiration (1500 mm) was measured by pan evaporation method according to Kader et al. (2017). Before sowing, loess-like loamy soil in experimental fields had gravimetric field water holding capacity of 25.6 mm and a mean bulk density of 1.38 Mg m−3. The soil (upper 40 cm layer) had a permanent wilting point of 6.7%. The soil physicochemical properties were estimated (Table 1).

Table 1 Soil physical and chemical properties in the experimental field
Full size table

2.2 Experimental Design

In a randomized complete block design (RCBD), there were ten treatments of three ridge-mulching materials with three ridge widths and flat planting (FP) as control with three replications each. Alfalfa was spaced 20 cm apart in experimental plots, which were 3 m wide and 10 m long. The ridge was 20 cm in height and sloped at an angle of 40° (Fig. 1). Manually compacted soil (CS), biodegradable film (BF), and plastic film (PF) as ridge mulching materials with three ridge widths (30, 45, and 60 cm) were studied. The ridges were covered with plastic and biodegradable films with edges buried 3–5 cm in the soil along the bases of the ridges. During ridge banking, manually compacted soil was created by hand, and after a few rain events, it became crusted. The plastic and biodegradable films (composed of starch) were 0.008 mm thick and were obtained from Shijiazhuang Yongsheng Plastic Plant Co. Ltd (China) and BASF Co. Ltd, (Germany), respectively.

Fig. 1
figure 1

Schematic diagram showing alfalfa production in ridge-furrow rainwater harvesting. FP, CS, BF, and PF were abbreviated for flat planting, ridges with manually compacted soil, mulched with bio-degradable film, and plastic film, respectively

Full size image

2.3 Field Management

After collecting litter and debris, plots on the fields were prepared and established in 2012. About 20–30-cm-thick surface layer of highly fertile soil was excavated and stacked. Using a slope meter and a tape measure, plots were manually constructed by molding soil surface into furrows and ridges at 40° slope to acceptable sizes (3.0–4.2 m wide and 10 m long, with 4 ridges and 3 furrows, except for controls) along breadths of plots. After plots were established, piles of dug-up soils were then spread uniformly over plots using a spade. The furrows were plowed, leveled, and harrowed for planting. Biodegradable films were placed on ridges and sides, and bases were buried in the soil to depths of 3–5 cm. This procedure was repeated on ridges in 2013, 2014, 2015, and 2016.

An alfalfa cultivar (No. 3 Gannong) bred at Gansu Agricultural University was hand sown at 22.5 kg ha−1 in 2012. This cultivar was selected and developed for its attractive characteristics such as improved drought, pests, and disease resistance. At depths of 2–3 cm, four rows were sown in a 60-cm-deep furrow with 20-cm gaps between two rows (Fig. 1). The ridge-furrow rainwater harvesting plots had 10 furrows with 40 rows of alfalfa, while flat planting field had 66 rows of alfalfa. Hand weeding was done throughout alfalfa cultivation seasons. During weeding, care was taken to avoid damaging the ridges. Neither fertilizer nor irrigation was applied to fields since alfalfa is a nitrogen-fixing legume and very sensitive to excessive soil water. The study was conducted under rain-fed conditions and climate. Field management and weed, insect, and disease control were all vigorously undertaken, as these activities have a significant impact on alfalfa production and quality of fodder. Harvest management entails determining number of cuts per season, date of cut, maturity stage, and time between cuts. During initial flowering phase and senescence, alfalfa was manually cut in all plots twice in 2012 and three times in 2013, 2014, 2015, and 2016 with sickles close to the soil surface.

2.4 Sampling and Measurements

In all experimental plots, soil samples were taken from 0 to 20, 20 to 40, and 40 to 60 cm and subsequently at 10-cm intervals at 300 cm soil layer with a soil auger to measure gravimetric soil water content (SWC). Readings on soil water content and moisture were recorded 24 h after every rainfall surpassed 5 mm (Hu et al. 2020). After undisturbed soil samples were dried in an oven at 105 °C for 24 h, bulk density (ρ) for 0–20 cm soil depth was determined (Verheijen et al. 2019). Alfalfa fodder samples (0.6 m × 0.6 m) were harvested from ridge-furrow rainwater harvesting and flat planting plots and weighed in kilograms. The weight of sample was then multiplied by total area of plot divided by area of sample collection. The value obtained was then converted to kilograms per hectare (0.0001 kg m−2 = 1 kg ha−1) to determine fodder yield. Approximately 1 kg of dried fodder sample was ground in a rotary mill to move through a 1-mm laboratory inspection screen to assess consistency of alfalfa fodder, as previously described by Wang et al. (2019). The Kjeldahl method was used to estimate total nitrogen (Total N) (Sebnie et al. 2020), and crude protein (CP) was calculated by multiplying nitrogen (%) by the constant factor of 6.25 to convert nitrogen values to crude protein (CP) (AOAC 2005; Rodrigues et al. 2018). Acid detergent fiber (ADF) and neutral detergent fiber (NDF) were evaluated using procedures by Grzegorczyk et al. (2017).

2.5 Crop Water Consumption, Water Use Efficiency, Soil Desiccation Index, and Cost–Benefit Analysis

2.5.1 Crop Water Consumption and Water Use Efficiency

Using the modified water balance formula developed by Mo et al. (2017), total actual crop water consumption (CWC, mm) and water use efficiency (WUE (kg ha−1 mm−1)) of alfalfa fodder yield for the growing seasons were calculated as

$$\mathrm{CWC}=\mathrm{P}+\left({\mathrm{W}}_{1}-{\mathrm{W}}_{2}\right)$$
(1)
$$\mathrm{WUE}=\mathrm{Y}/\mathrm{CWC}$$
(2)

where P is total alfalfa growing season precipitation (mm), W1 (mm) and W2 (mm) are amounts of soil water storage measured 1 day before sowing and after last cutting for all treatments, and Y is the fodder yield (kg ha−1). In addition, percolation and groundwater recharge are almost non-existent in this region (Ilstedt et al. 2016). Soil water storage in soil layers (W1 and W2) was calculated according to Mo et al. (2017) as:

$$\mathrm{W}=\left(\uptheta /100\right)\times\uprho \times \mathrm{h}$$
(3)

where θ is the soil water content (% kg kg−1), ρ the bulk density (g cm−3), and h is soil layer thickness (mm).

The soil water storage deficit (Da, mm) and degree of soil water deficit (DSW, %) were estimated as follows (Sun et al. 2018):

$$\mathrm{Da}=\mathrm{Fc}-\mathrm{SWS}$$
(4)

where Fc is field capacity (mm) and SWS is soil water storage (mm) in 300 cm depth.

$$\mathrm{DSW}=\left(\mathrm{Da}/\mathrm{Fc}\right)\times 100$$
(5)

Soil water depletion was calculated using the formula by Jin et al. (2019):

$$\Delta \mathrm{W}={\mathrm{SWS}}_{\mathrm{final}}-{\mathrm{SWS}}_{\mathrm{initial}}$$
(6)

SWSinitial is soil water storage (mm) before green-up, and SWSfinal is soil water storage (mm) after last cutting.

2.5.2 Soil Desiccation Index

The quantity of soil water content or moisture held in the soil for 2–3 days after surplus water has been purged away, and rate of decreasing water after a rainfall event is termed as field capacity (Fc) (Rai et al. 2017). Hence, in this study, maximum SWC in a growing year for studied profiles was assumed to be equivalent to volumetric Fc (θFc) of the year. Therefore, to obtain a more accurate and quantitative expression of the degree of soil water deficit, SDI was calculated following (Wang and Wang 2018):

$$\mathrm{SDI}=\left(\frac{{\mathrm\theta}_{\mathrm{Fc}}-{\mathrm\theta}_{\mathrm o}}{{\mathrm\theta}_{\mathrm{Fc}}-{\mathrm\theta}_{\mathrm{pwp}}}\right)\times100\left(\%\right)$$
(7)

where SDI is soil desiccation index and represents degree of soil water deficit, θFc is volumetric field capacity, θo is current SWC, and θpwp is SWC at permanent wilting point. The intensity of soil desiccation was segmented into four degrees based on values of calculated soil desiccation indices: severe soil water deficit, when SDI value is > 50%; moderate soil water deficit, when SDI value is 25–50%; minor soil water deficit, when SDI value is 0–25%; and no deficit, when SDI value is < 0 (Jin et al. 2019).

2.5.3 Cost–Benefit Analysis

The cost–benefit analysis included an assessment of total costs, income from hay sales, and net economic benefit (NEB). The total costs included cost of seed, biodegradable film, and plastic film. Income here refers to income from fodder yield. However, these estimates did not take into account fixed costs, such as value of land, interest on capital, or depreciation. Labor was self-provided at zero cost for ridging, cross-ties, weeding, application of biodegradable and plastic film mulches, and other sampling operations. The major output considered in this analysis was alfalfa fodder yield. The NEB was calculated by subtracting input cost from fodder yield income (Guo et al. 2019).

2.5.4 Statistical Analysis

Summary statistics of means (θ), standard deviations (SDs), ANOVA, and effects of treatments were computed using one-way Duncan’s procedure in SPSS software (version 26, IBM Corp., Chicago, IL, USA).

3 Results

3.1 Air Temperature and Rainfall

During alfalfa cultivation seasons (April–October), air temperatures were 21.2 (2012), 21.6 (2013), 21.1 (2014), 22.0 (2015), and 20.7 °C (2016), respectively. Compared to reference mean per annum temperature of 14.2 ℃, yearly air temperature was 13.9 (2012), 15.7 (2013), 14.9 (2014), 16.1 (2015), and 16.6 ℃ (2016) (Fig. 2a). This depicts a slightly higher air temperature during the alfalfa growing seasons.

Fig. 2
figure 2

Panel (a) shows mean monthly air temperature and panel (b) shows precipitation compared with the mean 46-year records (1970–2016) at the experiment site from 2012 to 2016

Full size image

Rainfall data showed a declining trend over the study period. Mean rainfall was 478.9 (2012), 492.8 (2013), 457.2 (2014), 298.2 (2015), and 311.95 mm (2016), while total rainfall (April–October) was 442.6 mm (2012), 466.2 mm (2013), 430.8 mm (2014), 262.3 mm (2015), and 288.85 mm (2016), respectively (Fig. 2b). Compared to the reference mean annum rainfall records of 385.3 mm, the mean rainfall for 5-year alfalfa cultivation seasons was 359.3 mm. This indicates a slightly decreasing trend in rainfall over the growing periods of alfalfa.

Rainfall in CS30 (3.27 and 4.94) was the highest as PF60 (0.04 and 1.30) had the lowest threshold rainfall in 2012 and 2016, respectively, when compared to flat planting. In 2013 and 2014, however, the highest threshold rainfall was recorded in CS30 (6.83 and 3.70) as the lowest was observed in BF60 (1.11 and 0.02), respectively. Furthermore, the highest threshold rainfall was recorded in CS45 (5.30), as PF60 (0.75) produced the lowest threshold rainfall when compared to flat planting treatment in 2015. Consequently, the proportion of variation between runoff and rainfall from 2012 to 2016 was the highest in PF with R2 = 1.00 and the lowest in CS with R2 = 0.56 (Fig. 3a). This suggests that PF treatment has an advantage in flood prevention which might ultimately lead to runoff compared to CS, BF, and flat planting treatments. This indicates the effectiveness of plastic film mulching in ridge-furrow rainwater harvesting system in conserving water under rain-fed agriculture.

Fig. 3
figure 3

Panels (a)–(c), (d)–(f), (g)–(i), (j)–(l), and (m)–(o) show rainfall-runoff relationship during alfalfa growing season in 2012, 2013, 2014, 2015, and 2016, respectively, for ridges with manually compacted soil, mulched with bio-degradable film, and plastic film. The ridge widths of CS30, CS45, and CS60 (BF30, BF45 and BF60 or PF30, PF45 and PF60) were 30, 45, and 60 cm, respectively, with 60 cm furrow width for all treatments

Full size image

3.2 Crop Water Consumption

Crop water consumption in CS45, CS60, and PF60 deceased during alfalfa cultivation in 2012, 2014, and 2015, respectively. In 2013, crop water consumption was greatly reduced in PF60 as compared to CS45, while in 2016; BF30 had the highest increase in CWC compared to BF60 which had the lowest increase in CWC. In 2012, mean CWC for FP was significantly higher than that for BF (mean of BF30, BF45, and BF60), and no significant (p > 0.05) differences were found between CS and FP. In 2013, CWC in FP was significantly higher than in PF, and no significant (p > 0.05) differences were established between BF and PF. Contrarily, in 2014, CWC in BF was significantly (p < 0.05) higher than that in FP, and no significant differences were found between BF and PF. In 2015 and 2016, CWC in BF was significantly (p < 0.05) higher than that in FP (Table 2). Consequently, over the 5 years of alfalfa cultivation, CWC decreased by 2.35% on average, ranging from 419.16 to 436.98 mm relative to flat planting. This implies that plastic film mulching in ridge-furrow rainwater harvesting system is effective in reducing soil water loss in rain-fed agriculture.

Table 2 Crop water consumption (mm) of alfalfa production in ridge-furrow rainwater harvesting from 2012 to 2016
Full size table

3.3 Soil Desiccation

Compared to CS and PF, soil water deficit (Da) was higher in FP and was lower in BF in 2012. This led to FP recording significantly (p < 0.05) higher mean soil desiccation compared to BF (mean of BF30, BF45, and BF60). Additionally, soil desiccation from 2013 to 2016 was significantly higher in FP than in PF (mean of PF30, PF45, and PF60). Consequently, moderate soil desiccation was found in CS (43.03%) while PF60 had the lowest soil water deficit (12.42 mm). The soil water depletion (− 0.49 m3 m−3) recorded in CS led to severe soil desiccation of 53.54% in 2013. Mean soil water depletion (ΔW) decreased by 0.54 in BF treatments compared to FP (Fig. 4b, d, e), suggesting moderate desiccation in BF treatment. Conversely, FP significantly increased the degree of soil water deficit (DSW) by 60.22% indicating severe desiccation while DSW was impacted by PF with moderate desiccation of 49.91% (Fig. 4b, c, d, e). This confirms that plastic film mulching in ridge-furrow rainwater harvesting reduces soil desiccation in alfalfa production under rain-fed conditions in semiarid areas.

Fig. 4
figure 4

Panels (a)–(e) show soil profile distribution of soil desiccation in 2012, 2013, 2014, 2015, and 2016, respectively, after the last cutting. FP, CS, BF, and PF were abbreviated for flat planting, ridges with manually compacted soil, mulched with biodegradable mulch film, and plastic film, respectively. The ridge widths of CS30, CS45 and CS60 (BF30, BF45 and BF60 or PF30, PF45, and PF60) were 30, 45, and 60 cm, respectively, with 60 cm furrow width for all treatments

Full size image

3.4 Fodder Yield and Water Use Efficiency

Alfalfa fodder yield increased gradually from April to June 2012 and from July to September 2012; fodder yield increased progressively after sowing in both ridge–furrow rainwater harvesting (RFRH) and FP. This trend occurred in subsequent years with yields in RFRH treatments substantially higher than those in FP. In 2012 and 2013, yield decreased in BF60 (2.02%) and CS30 (0.83%) as compared to increase in yield in PF45 (16.05%) and PF30 (43.4%), respectively. Conversely, in 2014 and 2015, yield was decreased by 4.8 and 9.1% in CS45 and increased by 64.11 and 62.7% in PF30 relative to FP, respectively. In 2016, however, yield increased in PF45 by 22.3% and decreased in CS60 by 1.5% compared to FP (Table 3). On average, CS had no increase in yield of alfalfa over the 5 years growing period, while BF and PF had a 28% and 33% yield increase compared to FP. In addition, average mean yield increase over the 5 years ranked PF > BF > FP > CS. This implies that, averagely, plastic film mulched ridge-furrow rainwater harvesting with 30 cm ridge width is efficient in increasing alfalfa fodder yield under rain-fed conditions.

Table 3 Fodder yield (kg ha−1) and water use efficiency (WUE) (kg ha−1 mm−1) of alfalfa in ridge-furrow rainwater harvesting from 2012 to 2016
Full size table

Conversely, water use efficiency (WUE) decreased in CS30 in 2012 and 2013 and correspondingly increased in PF45 and PF30 (Table 3). Again, in 2014 and 2015, PF30 had the highest increase in WUE while CS45 and CS60 had a decrease in WUE. However, in 2016, WUE was increased in PF45 but decreased in CS45 during the alfalfa growing seasons. Over the 5 years of alfalfa cultivation, WUE increased by 19.79% on average and was highest in PF30 by 28.19 kg ha−1 mm−1 relative to flat planting. This suggests that plastic film mulched ridge-furrow rainwater harvesting with 30 cm ridge width is efficient in increasing alfalfa water use efficiency under rain-fed conditions.

3.5 Fodder Quality

We assessed how the treatments affected nutrient compositions of alfalfa fodder by analyzing the fodder quality. Crude protein (CP) in PF60 was 196 g kg−1 and 214 g kg−1 in 2012 and 2015, respectively, while in 2013, 2014, and 2016, CP was 217, 215, and 189 g kg−1 in BF60, respectively (Table 4). This indicates a significant increase in PF (28.77%; p < 0.05) relative to FP, CS, and BF. In terms of acid detergent fiber (ADF), flat planting (FP) had 394 g kg−1 ADF as the highest, and BF treatments had the lowest mean ADF of 348 g kg−1, with BF60 recording 345 g kg−1 mean ADF. This implies that flat planting increased ADF during the alfalfa growing seasons. A similar trend was found for neutral detergent fiber (NDF). Flat planting had the highest NDF (505 g kg−1) while BF treatments had the lowest mean NDF of 471 g kg−1 with BF45 and BF60 recording 466 g kg−1 each. Averagely, among all treatments of alfalfa cultivation over 5 years, CP ranged from 169 to 202 g kg−1, ADF ranged from 338 to 394 g kg−1, and NDF ranged from 464 to 505 g kg−1. As demonstrated by this study, CP was significantly (p < 0.05) higher in PF treatments with FP recording the highest mean ADF and NDF compared to CS and BF. This suggests that plastic film mulched ridge-furrow rainwater harvesting with 60 cm ridge width is efficient in increasing alfalfa fodder crude protein content under rain-fed conditions.

Table 4 Crude protein (CP) (g kg−1), neutral detergent fiber (NDF) (g kg−1), and acid detergent fiber (ADF) (g kg−1) of alfalfa in ridge-furrow rainwater harvesting from 2012 to 2016
Full size table

3.6 Cost–Benefit Analysis

Smallholder farmers have less income to adopt two ridge widths to determine which is more beneficial to them; hence, we conducted an economic benefit analysis as different mulching materials were used with variances in input costs of the mulching materials (Table 5). The average input cost (US Dollar (US$)) over 5 years was rated as follows: BF > PF > CS > FP. The average input cost values under BF and PF were 245.16 and US$148.87 ha−1, respectively, higher than CS and FP. The plots’ most important output value was fodder yield, which had a consistent market price (local price) between and within experimental seasons. In this study, output values from 2012 to 2016 for various treatments were ranked as PF > BF > FP > CS per net income. The PF60 treatment registered the highest net economic benefit (NEB) throughout 5 years of alfalfa cultivation. This indicates that an investment in PF treatment with a 60 cm ridge width by a smallholder farmer is technically and economically viable.

Table 5 Cost–benefit analysis (USD ha−1) of alfalfa production in ridge-furrow rainwater harvesting from 2012 to 2016
Full size table

4 Discussion

When rainfall becomes the only recorded climatic parameter, runoff, a rare but vital parameter for soil and water conservation research, becomes difficult to measure. Due to interception and initial high infiltration losses, a threshold rainfall is always required before runoff occurs. The threshold rainfall is the minimum amount of rainfall above which runoff occurs, calculated by plotting daily rainfall depth against corresponding runoff depth and fitting it with a least-square curve (Liu et al. 2020a). In this study, all ridge-furrow rainwater harvesting treatments had low threshold rainfalls for runoff, though threshold values increased slightly in ridges compacted with soil. Other studies in northern Ethiopia (He et al. 2018b; Li et al. 2017c; Milkias et al. 2018) corroborated findings of this study. Threshold rainfalls for previous studies range from 5 to 8 mm for short enclosures and 3 mm for grazing lands (Dao et al. 2020). Xin et al. (2021) reported a 2-mm rainfall threshold for cultivated land and ascribed it to lower infiltration capacity of soils in semiarid environments. The slight increase in rainfall thresholds in ridges compacted with soil is most likely attributable to water ponding in ridges and improved soil infiltration capacity due to reduced ridge widths.

Variability in amount and distribution of seasonal precipitation, which involves evaporation from soil surface and crop transpiration, is a major source of variation in CWC on the Loess Plateau (Do and Yin 2018). We recorded a 2.35% decrease in CWC in the ridge-furrow rainwater harvesting system with mulching over 5 years of alfalfa cultivation compared to flat planting. This is consistent with Zhao et al. (2018), who reported simulation of heat flow and soil water in RF with plastic film mulching, decreased CWC on Chinese Loess Plateau where plastic film mulching was less effective for increasing WUE in rain-fed agriculture. Furthermore, Dang et al. (2016) in 2015 reported 56.2% higher yield under plastic film-mulched ridge-furrow (RF), 63.4% higher WUE, and 15.0% lower water use (CWC) than FP, respectively. Plastic film mulching with RFRH markedly increases WUE and improves crop production as a potential soil amendment for sustainable rain-fed agriculture (Ding et al. 2019).

We assessed the effect of ridge-furrow rainwater harvesting treatments on alfalfa yield and soil desiccation over the 5 years. Mean soil desiccation in this study compared to FP decreased by 0.54 in BF treatment. This may be attributed to high weed control potential of biodegradable films (Marí et al. 2019), which decreases bulk density, hence increasing WUE and crop production (Gu et al. 2020). Additionally, Sekara et al. (2019) and Caruso et al. (2019) reported biodegradable films differentially influenced soil temperature and humidity. This could be beneficial in arid and semi-arid areas characterized by high temperatures that could damage crops. In this study, optimum ridge width in RFRH system to increase yield, improve WUE, and thus reduce soil desiccation was 60 cm for alfalfa production in northwestern China. Hu et al. (2020) reported that ultimate ridge and furrow width for increasing WUE in alfalfa production was 60 cm, thus corroborating our finding. Furthermore, Luo et al. (2021) also reported a ridge width of 0.3 m and a furrow width of 0.6 m for use in semi-arid areas for alfalfa production.

There are two technical components of RFRH system: rainwater harvesting and mulching. The system improves soil water content and soil temperature, prolongs period of moisture availability, assists in weed control, and enhances agricultural production (Zhang et al. 2019). Simultaneously, this technique uses contour ridges, which are intended to reduce runoff and soil erosion (Li et al. 2021). Alfalfa yield was increased by 33% in this study with plastic film, which could be due to lower crop water consumption in mulched fields. This is consistent with Gu et al. (2018), who reported a range of 10.7–40.3% increase in alfalfa yield in PF treatments and 14.2–28.3% decrease in alfalfa yield in CS treatments. There were higher and comparable yields in 2015 and 2016 under less rainfall conditions compared to yields recorded in 2013 and 2014. This may be attributed to the high-yielding ability of the alfalfa cultivar. In addition, continued field maintenance was undertaken in years following establishment to achieve high yields and sustain a desirable level of production over time. Furthermore, alfalfa is exceptionally drought resistant and does not require additional water to produce higher yields (Liu et al. 2018b). Alfalfa roots typically grow to a depth of 3–5 m and can reach up to 8–15 m in soils. Consequently, when surface water is scarce, alfalfa may depend on moisture stored deep in soil profiles (Jia et al. 2020). Furthermore, alfalfa’s ability to use residual winter rainfall makes it possible to grow for 4–8 years and rapidly in warm conditions in spring. This is in contrast to summer-grown annual crops, which must be replanted every year (water use efficiency is low during this time) (Mak-Mensah et al. 2021c). Alfalfa is also valuable due to its deep rooting ability to exploit conserved soil moisture, ability to withstand droughts, and ability to produce high yields in drought conditions (Gruffat et al. 2020).

We evaluated treatment-specific effect of ridge-furrow rainwater harvesting on WUE and fodder quality over the 5 years. WUE correlated with fodder and grain yield per unit of alfalfa crop water consumption. In comparison with flat planting, plastic film mulched treatment increased mean WUE by 35% during alfalfa cultivation seasons. This finding is corroborated by Berhanu et al. (2020), who reported that WUE of long-term crops increased from 0.23 to 0.90 kg m−3, due to improved soil management practices. Our finding also compares well with reported increase in wheat yield and WUE from 4422 irrigated sites in 22 provinces of China (Yoon and Choi 2020). However, locust and poplar (Populus tremuloides) trees had mean WUE of 0.74 kg m−3 and 0.67, respectively (Dornbush and von Haden 2017). Therefore, improving WUE in rain-fed agriculture is crucial since there are high demands for food production in China. Water-saving practices have only just begun to be adopted by farmers, and adoption has been poor, which may be ascribed to poor sensitization and extension services (Bhatt et al. 2021). In semi-arid areas, crop improvement programs are aimed at increasing yields (Kong et al. 2020). Therefore, RFRH system is an effective method for low-income farmers to improve rainwater use efficiency in semiarid areas (Wang et al. 2021). In addition, co-application of plastic film mulched RF with biochar may potentially reduce negative effects of plastic film application, such as greenhouse gas emissions and soil residue accumulation (Mak-Mensah et al. 2021a).

Fodder quality is a reflection of essential nutrient elements available to animals for their daily nutrient requirements (Hakl et al. 2017). Thus, suitable fodder quality is critical for high milk and meat production. Fodder quality analysis can help with fodder processing, animal feeding, and nutrition. Optimal fodder quality occurs before plant maturity and seed production (Masikati et al. 2017). Although it is desirable in agricultural production to achieve the highest possible fodder yield, it is also important to have high fodder quality (Sandhu et al. 2020). Crude protein concentration was higher in the leaves than in stems, while ADF and NDF concentrations were higher in stems than in leaves (Hofmann et al. 2019). In this study, CP was high in PF treatments; while FP had the highest mean ADF and NDFs compared to CS and BF (p < 0.05). High soil water content in RFRH system decreased fodder stem-leaf ratio, resulting in higher CP content and lower ADF and NDF contents compared to FP. The RFRH system may increase alfalfa fodder quality as plant maturity is delayed, and production of cell wall components is reduced (Sandhu et al. 2020). Drought stress harms fodder quality, accelerating decline in CP content. The RFRH systems, especially PF, increased soil water content and resulted in high CP content and low ADF and NDF contents. Alfalfa and other legumes have distinct leaves and stems, whereas leaf and stems are intertwined in grasses. Leaves contain more digestible nutrients in contrast with stems, which reduces fodder quality in alfalfa (Rezaeian et al. 2020).

Since smallholder farmers often search for alternative methods to increase productivity and reverse land degradation, biodegradable and plastic film mulches could be potential materials for improving yield while increasing provision of other important ecosystem services. A cost–benefit analysis (CBA) was performed to evaluate profitability of application of these materials. The CBA clearly shows varying costs and benefits associated with application of biodegradable and plastic film mulches in alfalfa production over the 5 years. In this study, plastic film mulch yielded the highest cost–benefit ratio compared to FP and CS with no marked difference between them. This is consistent with Ma et al. (2018), who reported positive impact of PF on NEB in spring maize, suggesting that northwestern China could be a maize belt if supported by widespread adoption of PF. Additionally, Ma et al. (2018) noted that use of a biodegradable film may be an option to achieve economic benefit and minimize plastic film pollution. Furthermore, Liu et al. (2020) reported that using 60-cm plastic-mulched ridges increases high crop yield and cost–benefit ratio.

5 Conclusions

Mulch application on semiarid lands can be a viable strategy to strengthen the ridge-furrow rainwater harvesting systems under rain-fed agriculture. Results revealed that integration of ridge-furrow rainwater harvesting and mulching in alfalfa production over the 5 years increased alfalfa fodder yield and quality with higher cost–benefit ratio compared to flat planting. As ridge-furrow rainwater harvesting systems with mulching could be employed for prevention of runoff and soil erosion through allocation of water resources, effective designs should accommodate an adaptive structure for management.