JlA Li-guo, CHEN Yang, QlN Yong-lin, LlANG Rui-fang, CUl Shi-xin, MA Zhong, FAN Ming-shou

1 College of Agronomy, Inner Mongolia Agricultural University, Hohhot 010018, P.R.China

2 College of Ecology & Environmental Science, Inner Mongolia Agricultural University, Hohhot 010019, P.R.China

3 Department of Biology, Truman State University, Kirksville, MO 63501, USA

Abstract Yin-mountain Hilly Area is one of the ideal regions for potato (Solanum tuberosum) production in China. However, potato yield is severely limited as a result of rain-fed crop mode due to water deficiency, as well as an inadequate farming practices.In this study, yield gaps were determined by using attainable yield (Ya) as a benchmark under optimized management practices, i.e., micro-ridge and side planting with plastic-mulching (MS), and flat planting with plastic-mulching (PM). The yields under MS and PM modes are defined as Ya1 and Ya2, respectively. Under the same field with MS and PM modes but different densities and fertilizer usages and so on, it was defined as simulated farmers’ practices. The yield of simulated farmers’ practices (Yf1) reached 57.3 and 69.6% of Ya1 and Ya2, respectively, while the average yield of 298 randomly surveyed farmers (Yf2) reached only 37.0 and 47.8% of Ya1 and Ya2 for rain-fed potato, respectively. The gaps of water use efficiency exhibited similar pattern. Further analysis shows that improper measures in rainwater conservation and accumulation, and other management practices contributed to 18.5, 18.2, and 42.6% of yield gap between Ya1 and Yf2.Improper nutrition management, including overuse of nitrogen and the deficiency of phosphorus and potassium supplication,was one of the important reasons of yield gap. The results indicate the possibilities of increasing rain-fed potato yields by optimized water and fertilizer managements in the Yin-mountain Hilly Area.

Keywords: rain-fed, potato, yield gap, water management, Yin-mountain Hilly Area

1. lntroduction

Increase in crop production is necessary to keep pace with continual rise in food demand driven by population growth(Ittersum et al. 2013). China has a large population and is expected to reach 1.5 billion in 2033. Since arable land resources are limited and decreasing, it will be necessary to increase crop production by at least 35% during the next 20 years (Zhang 2011). Potato is the fourth largest crop in China following rice, maize, and wheat, ranking No. 1 in both yield and planting area in the world. In 2015,China launched the potato staple food strategy in order to ensure food security at the country level (Fan et al. 2015).Thus, potato crop is expected to develop rapidly and play increasingly important roles in the near future.

The yield gap has been used to dissect major limiting factors in production, and also as an indicator for the possibility of increasing crop yields in a given region in many recent studies (Lobell et al. 2009; Finger 2011; Hall and Richards 2013; Meng et al. 2013; An et al. 2015; Svubure et al. 2015). Determining yield gaps for specific crops is beneficial to prioritizing research and adopting practices to reduce these gaps (Abeledo et al. 2008; Tittonell et al.2008). However, it is hard to eliminate yield gaps completely due to difficulties in controlling the many factors affecting yield in actual farming practices.

Ascertaining yield potential is the first step toward yield gap analyses. Yield potential usually refers to the yield of a crop cultivar grown with sufficient supply of water and nutrients, and with biotic stress effectively controlled (Evans 1993; Van Ittersum et al. 2013). In addition, attainable yield is also used to estimate yield gap from the average farmers’yield in some research (Hall et al. 2013; Pasuquin et al.2014; Tejendra and Allen 2015). Attainable yield is a contextdependent variable that is affected by environmental,economic, and social factors. Compared to yield potential,attainable yield is a more appropriate benchmark in yieldgap analysis for a specific region, therefore can be used to provide a strategy for decreasing yield gap.

The results from Svubure et al. (2015) showed that actual farmers’ yield of potato ranged from 8-35% of the potential yield estimated by LINTUL-POTATO model, translating into a yield gap of 65-92% in Zimbabwe. Another report from Chile demonstrated that the actual mean yield of farmers investigated was 31 t ha-1, and the potential yield was on average 74 t ha-1in potato, i.e., less than half of the potential yield was achieved (Haverkort et al. 2014). It is clear that there is a huge potential to increase potato production across the world.

2. Materials and methods

2.1. On-farm trials

The experiment was conducted at a farm of Wuchuan County, in Inner Mongolia of China (41°15′N, 111°29′E) on a Chestnut soil (Chinese classification), typical for the Yinmountain Hilly Area during the potato growing season (May-September) of 2013 and 2014. The soil was sandy loam with electrical conductivity (EC) values of 180-230 μs cm-1, 18.7-21.1 g kg-1organic matter, 1.34 g kg-1total N, and available P (Olsen-P) and K (exchangeable K) of 14.2 and 128.3 mg kg-1, and the pH was 8.1-8.3. The farm is located at the Yin-mountain Hilly Area and the average annual precipitation was 354 mm with average evaporation of 2 068 mm in the past 30 years. The average photosynthetically active radiation (PAR) was approximately 1.83×106kJ m-2and the mean temperature was 14.2°C during potato growing season. The experimental fields differed from experimental years due to crop rotation. Spring wheat or sunflower was selected as preceding crop in the experiments.

The experimental treatments included conventional flat planting (FP), flat planting with plastic mulching (PM), and micro ridging with plastic mulching and side planting (MS).Conventional flat planting was designed based on local farmers’ practices, i.e., sowing without any cover and ridge,at a density of 37 500 hill ha-1with 0.5 m row spacing and 0.53 m hill spacing, treated with N (75 kg ha-1from urea)and P2O5(45 kg ha-1from single superphosphate), but no potassium fertilizer were applied during sowing.

In flat planting with plastic-mulching mode, transparent plastic film was used as cover during sowing, and two rows were laid out for each 1.2 m width film applied. Different from the FP treatment, a density of 52 500 hill ha-1with 0.5 m row spacing and 0.38 m hill spacing, N (105 kg ha-1as urea), P2O5(45 kg ha-1as single superphosphate), and K2O (90 kg ha-1as KCl) in PM treatment were applied during sowing stage.Micro-ridge and side planting with plastic-mulching was a novel cultivation mode on rain-fed potato developed by our research group based on local soil characteristics. In this mode, the soil was ridged to 12 cm above field level with ridges of 50 cm wide at its base, a 75-cm wide transparent plastic film was covered along the ridge. Before covering,fertilizers of the same type and quantities as PM treatment were applied. Seed potatoes were sowed with 0.5 m row spacing at the middle of the ridge slope, with the same density as that of the PM treatment (Fig. 1).

The experiments were laid out in a completely randomized block design with three replicates, each plot was 100 m2.The sowing and harvesting dates were 15 May and the end of September respectively, a short delay may happen due to weather condition. Potato cultivar Kexin 1, widely used in this region, was selected for all experiments. Disease,weeds, and pest control, as well as other managements were done by local standard methods.

2.2. Local farmer survey

Fig. 1 Schematic diagram of micro-ridging and side planting with plastic-mulching mode. Two ridges are one operating unit and two units are shown as two arrows with different directions.The space between two units is 300 mm and the space between two ridges is 100 mm. The ridge height is 120 mm.The distances among seedlings are 540 mm. Two rows seedling in each ridged sides are arranged alternately but the row spacing is not always equal, the distance in one ridge,between two ridges and between two units are 340, 260, and 460 mm, respectively.

Farmers’ practices on rain-fed potato were also investigated in 2013 and 2014. The sites-surveyed were along Yinmountain Hilly Area in Inner Mongolia Autonomous Region,including parts of a total of 13 counties in Hohhot, Baotou,and Wulanchabu. The farm surveys were conducted through face-to-face interviews with a total of 298 randomly selected farmers, including 121 in 2013 and 177 in 2014.Among these farmers, 42 households used film cover.Total tuber yield, planting area, sowing and harvesting time,preceding crop, fertilizer use, control of disease, weeds and pests on rain-fed potato only were recorded for each individual farmer.

2.3. Yield gap analyses

Yield gap (Yg) in this study was defined as the difference between attainable yield (Ya) and farmers’ yield (Yf) of rainfed potato (i.e., Yg=Ya-Yf, in t ha-1), and as percentage of the attainable yield (Yf/Ya×100%). Both attainable yield and farmers’ yield were based on fresh tuber weight. Attainable yield for rain-fed potato refers to the yield obtained when precipitation is the only water source but nutrients are fully supplied, pests and weeds are efficiently controlled, and farming technique and management are optimized. Two kinds of attainable yields were defined in this study, one is from novel rainfall accumulation mode (MS) and regarded as attainable yield 1 (Ya1), the other is from flat planting with plastic-mulching mode and regarded as attainable yield 2(Ya2). Actual yield is the yield achieved in a specific year under current production and management techniques at the farm. In this study, the yield based on local farmers’ practices on density and fertilizer application from experimental field was defined as farmers’ yield 1 (Yf1), while actual yield based on local farming practices from the survey was defined as farmers’ yield 2 (Yf2), of which the farmers’ yield using film cover was defined as Yf2.1, and farmers’ yield without using film cover as Yf2.2. Ya1, Ya2, Yf1, and Yf2 were the average yield of 2013 and 2014.

Yield gap between Ya1 and Yf1 was calculated during Ya1-Yf1 for 2013 and 2014, and then averaged; the same method was used for gap calculation between Ya2 and Yf1,Ya1 and Ya2. The percentages of the attainable yield were calculated using corresponding data.

Gretel flew straight to Hansel, opened the little stable-door, and cried: Hansel, we are free; the old witch is dead. Then Hansel sprang like a bird out of a cage when the door is opened.52 How they rejoiced, and fell on each other s necks, and jumped for joy, and kissed one another!

2.4. Water use efficiency (WUE) gap

WUE was expressed as the fresh tuber yield per hectare and unit precipitation during potato growing season (kg mm-1ha-1) in this study. Attainable water use efficiency(WUEa), therefore, refers to WUE in obtaining attainable yield; correspondingly, WUEf is the WUE of actual farmers,and WUEa1, WUEa2, WUEf1, and WUEf2 are the water use efficiency for yields of Ya1, Ya2, Yf1, and Yf2, respectively.Water use efficiency gap (WUEg) in this study was defined as the difference between WUEa and farmers’ water use efficiency (WUEf) of rain-fed potato, expressed as kg mm-1ha-1(WUEg=WUEa-WUEf), and as percentage of the attainable water use efficiency (WUEg (%)=WUEf/WUEa×100).

2.5. Statistical analysis of data

All statistical analysis was performed using the SPSS software package (SPSS 13.0, SPSS Inc.). Means of potato tuber yield and regression slopes of yield gap in response to precipitation during seedling growth and development were compared at a 0.05 level of significance.

3. Results

3.1. Attainable yields and farmers’ yields

During the 2013 and 2014 potato growing period, average precipitation was 259 mm. When fertilizers were fully supplied with pests, diseases, and weeds controlled,attainable fresh tuber yield of rain-fed potato in Yin-mountain Hilly Area was 22.2 t ha-1(Ya2) using plastic film as cover to keep soil water (PM mode). When rainfall accumulation was further optimized by combining ridge and film-cover(MS mode) under above management, attainable fresh tuber yield (Ya1) increased by 13% to 25.1 t ha-1(Fig. 2-A).

The simulated farmer’s yield (Yf1) was 14.6 t ha-1averaged across 2013 and 2014 (Fig. 2-A), which was significantly lower than Ya1 and Ya2. The mean value of the actual farmers’ yield (Yf2) from a survey of 298 farms in 2013 and 2014 was 9.3 t ha-1(Fig. 3-A). The average tuber yield was 10.8 t ha-1(Yf2.1) on these farms using film cover, but only 7.9 t ha-1(Yf2.2) on farms not using film cover (Fig. 4-A).

3.2. Yield gaps

The average yield of rain-fed potato from simulated farmer practice (Yf1) was lower than the experimentally attainable yields Ya1 and Ya2 by 11.5 and 6.4 t ha-1(Fig. 2-B), and this represented 57.3 and 69.6% of attainable yields (Fig. 2-C).In comparison, larger yield gaps of 15.8 and 12.9 t ha-1were exhibited between attainable yields (Ya1, Ya2) and yield from local farmer practices (Yf2) (Fig. 3-B), which represented only 37.0% of Ya1 and 47.8% of Ya2 (Fig. 3-C).

The survey showed that very low percentage of local farmers fertilized at optimal amount scale. Only about 20% farmer applied fertilizer at the suitable level based on the nitrogen, phosphorus, and potassium fertilizer rates in local farming practice. Too much nitrogen and too less phosphorus and potassium fertilizer amount, both higher than 50% farmers, are the major problems on potato nutrition management (Table 1).

Fig. 2 Attainable yield and yield gaps of rain-fed potato under experimental field condition. A, attainable yields of rain-fed potato in micro-ridge with plastic mulching and side planting system (Ya1) and flat planting with plastic-mulching (Ya2), yield of simulated farmer practice (Yf1). B, yield gap between Yf1 and Ya1, Ya2, respectively. C, the percentage of Yf1 as Ya (Yf1/Ya) includes Yf1/Ya1 and Yf1/Ya2. Values are means of 2013 and 2014 for Ya1, Ya2, and Yf1, respectively. Means denoted by the same letter did not significantly differ at P＜0.05 according to Duncan’s multiple range test. Bars are SD.

3.3. Temporal yield variability

Yield of rain-fed potato was strongly affected by in-season precipitation at the Yin-mountain Hilly Area. Average yield was 7.8 t ha-1under 246 mm precipitation in 2013, and 10.1 t ha-1under 272 mm precipitation in 2014, exhibiting closely positive correlation between tuber yield and precipitation during the growing period.

However, substantial variation in rain-fed potato yield(Yf2) within each year was observed, as well as range shift of actual yields between years (Fig. 5). Although mean values of field yields from individual farmers were lower than attainable yields (Ya1) in the two-year surveyed, yields in some fields exceeded the attainable yields in each year.Less than 5% of yields from surveyed farmers were higher than attainable yields, indicating that good practice by farmers can produce yield equal to or better than attainable yield.

3.4. Gaps of WUE

Fig. 3 Yield of rain-fed potato under local farming practice (Yf2) and its gaps with attainable yields (Ya1 and Ya2). A, average farmers’ yield of local rain-fed potato according to two-year survey data (2013 and 2014). Yield gap (B) and the percentage (C)of Yf as Ya (Yf/Ya) between Yf2 and Ya1, Ya2 respectively. Solid line and small square in box indicate median and mean yields,respectively. The box boundaries indicate upper and lower quartiles, the whisker caps indicate the 90th and 10th percentiles, and the diamonds indicate outliers.

Fig. 4 Component analyses of rain-fed potato yield gap. A, yields under local farming practice with (Yf2.1) and without (Yf2.2) plastic film cover through 2013-2014 survey. B, yield gaps between Yf1 and Yf2.2, Yf2.1 and Yf2.2, Ya1 and Ya2. C, the corresponding percentage as yield gap (Ya1-Yf2). Solid line and small square in box indicate median and mean yields, respectively. The box boundaries indicate upper and lower quartiles, the whisker caps indicate the 90th and 10th percentiles, and the diamonds indicate outliers.

WUE (kg mm-1ha-1) in this study refers to fresh tuber weight produced per hectare under each mm precipitation during potato growing season. WUEa was the efficiency under attainable yield. WUEa1 and WUEa2, i.e., the average water use efficiency when yields were Ya1 and Ya2 accordingly, reached 97.97 and 86.84 kg mm-1ha-1,respectively (Fig. 6-A). WUE of simulated farmer practices(WUEf1) and actual farmer practices (WUEf2) were 57.53 and 40.20 kg mm-1ha-1, respectively (Fig. 6-A).

The average WUEf1 was lower than the experimentally attainable WUEa1 by 40.44 kg mm-1ha-1, and WUEa2 by 29.31 kg mm-1ha-1(Fig. 6-B), in other words, achieving 58.72 and 66.25% of attainable WUE for rain-fed potato(Fig. 6-C). Consistent with the yield gaps, WUE gaps between local farmer practices and experimental practices leading to the attainable yields were larger, i.e., 57.77 kg mm-1ha-1for WUEa1-WUEf2, and 46.64 kg mm-1ha-1for WUEa2-WUEf2 (Fig. 6-B); in other words, WUE under local farmer practices was only 41.04% of WUEa1 and 46.30%of WUEa2 (Fig. 6-C).

4. Discussion

Water-limited potential yields have been used as a benchmark for yield gap estimation in some rain-fed crops(Berry and Spink 2006; Ittersum et al. 2013). Similar to the potential yield, it is generally estimated using crop growth model, which has well-developed in some major cereal crops such as rice, maize, and wheat for accurate calculation.However, similar models for potato are not yet developed enough to become well established (Travasso et al. 1996;Haverkort et al. 2015; Svubure et al. 2015). Therefore, in this study, we used attainable yield obtained under optimal experimental condition as a benchmark to analyze yield gaps for rain-fed potato in the Yin-mountain Hilly Area.

Table 1 Percentages (%) of nitrogen (N), phosphorus (P), and potassium (K) fertilizer rates in local farming practice

4.1. Yield gap resulted by water management

MS is a novel cultivation mode developed on the basis of local soil and precipitation characteristics, which has been verified with well rainwater collection effect in our recent study (Chen et al. 2017). Consistently, the potato yield under MS increased significantly over local farmer practices,which was defined the attainable yield (Ya1). Actually,enhancing rainwater collection by ridging and covering has been applied in some other crops (Wang et al. 2009; Han et al. 2013; Qin et al. 2014). Its effect on yield increase was obvious in potato through our two-year comparative trials (Fig. 2-A). Plastic film covering has been proved to be effective on soil water conservation but rarely used in local rain-fed potato production. Therefore, the yield of flat planting with PM method under optimal management practices is considered to be another attainable yield (Ya2).

In order to clarify the detail of local yield gap, the simulated farmer yield (Yf1) via field trail and the actual farmer yield survey were conducted. Such survey is always used for evaluating the actual farmer’s practices by collecting enough samples, which has been got good results in crops such as maize, soybean, and wheat (Wiese 1982;Lobell et al. 2005; Grassini et al. 2011, 2015; Villamil et al.2012). Our results show that the two types of attainable yields are higher than both the simulated farmer yield (Yf1)and the actual farmer yield (Yf2) (Figs. 2-A and 3-A). These gaps could be larger, although management practices in attainable yield experiments such as weed and disease were well controlled, they could be improved further. Thus,it is possible to increase the yield of local rain-fed potato by more than 15.8 t ha-1.

Fig. 5 Scattergram showing distribution of individual field yields for various locations in 2013-2014. Each diamond is a single data-point and diamonds are clustered around the vertical line for each year of record. Solid circle stands for the average farmer yield for each year. Arrow heads next to each vertical series indicate the attainable yield (Ya1) for that year.

Fig. 6 Attainable water use efficiency (WUEa), water use efficiency of farmer practices (WUEf) and efficiency gaps of rain-fed potato. A, attainable water use efficiencies of rain-fed potato in micro ridge with plastic mulching and side planting system (WUEa1)and flat planting with plastic-mulching (WUEa2), yield of simulated farmer practice (WUEf1) and actual farmer practice (WUEf2).B, water use efficiency gap between WUEa and WUEf (WUEa-f). C, the percentage of WUEf as WUEa (WUEf/WUEa). Solid line and small square in box indicate median and mean yields, respectively. The box boundaries indicate upper and lower quartiles,the whisker caps indicate the 90th and 10th percentiles, and the diamonds indicate outliers.

Because of the characteristics of local natural precipitation and sandy loam of soil, it is difficult to meet potato water requirement if no effective water management. Through conserving soil water by using plastic film covering the yield gap could decrease 17.0% according to the value of Yf2.1-Yf2.2, because film covering measurement is the major difference between Yf2.1 and Yf2.2 (Fig. 4). A little different from 33.9-92.5% yield increase reported,maybe due to the differences on climate and precipitation(Zhao et al. 2012). The gap decreased 2.87 t ha-1and the ratio was 16.8% by ridging based on the gap between Yf2 and Ya1 (Fig. 4-B and C). Since the only difference between PM and MS cultivation mode was ridging of field for rainfall accumulation, the yield gap indicates the gap caused by water accumulation. Since local rainfall is often not enough to meet normal water requirement of potato growth and development, any increase of soil water is definitely beneficial for potato yield formation. As mentioned above, the optimized rainfall management could decrease more than 1/3 yield gap, implying decreasing water use efficiency gap is one of most effective way to achieve yield potential of local rain-fed potato, and adopting techniques for accumulating and conserving more rainwater in soil is of great significance.

4.2. Yield gap resulted by fertilizer management

Another main factor affecting yield gap, and corresponding gap of water use efficiency, is unreasonable fertilization practices according to the survey results. Base on the gap between Yf1 and Yf2.2 which stands for gaps of local farmer on crop managements except for on water, the yield gap resulted from improper management is 39.2% and 6.7 t ha-1,including planting density, fertilizer application, disease,weeds and pest control, etc. (Fig. 4-B and C). Although many managements result in yield gap, improper fertilization is one of the most important contributors to the yield. Chen et al. (2012) reported that the optimal amounts are 120-180 kg N ha-1, 60-90 kg P2O5ha-1, and 105-150 kg K2O ha-1for local rainfed potato production. However, only about 20% farmers fertilized at this optimal level based on the local farmers survey (Table 1). Overuse of nitrogen fertilizer and deficiency of phosphorus and potassium fertilizer rates are another major cause for yield gap.

The two types of gaps do not add up to reach 100%attainable yield, suggesting “additive effect” might be exist due to the interaction between water managements and other practices. Indeed, there are large variations among different farmer on potato yield in specific year (Fig. 5).Farmers with lower yield might have adopted multiple improper farming practices, therefore the “additive effect”further decreased their yield. In contrast, farmers with relatively high yield likely used proper practices such as best fertilizer management. However, most farmers’ yields were far below the attainable yield. This was due to the combined effect of a lack of water management measures and other practices that were not optimized. Optimizing water and fertilizer managements are available ways to close yield gap for local rainfed potato production in the future.

5. Conclusion

Attainable yield as benchmark can be used for quantifying yield gaps in rain-fed potato production. Based on the attainable yield of water accumulating cultivation mode(MS), it was found that 36.7% yield gap could be decreased in local rain-fed potato production. Optimized fertilization rates, especially N, P, and K fertilizers, could reduce the most of another 42.6% yield gap of rain-fed potato in Yinmountain Hilly Area. Thus, optimizing rainwater and fertilizer management is the major way to increase local potato yield in the future.

Acknowledgements

This study was supported by the Special Industry Foundation of Ministry of Agriculture of China (201303104), the National Natural Science Foundation of China (31360502), and the China Postdoctoral Science Foundation (2015M572633XB).