1. Introduction
China is the fourth-largest soybean producer in the world after the United States, Brazil, and Argentina (Kou and Feng 2008, 4–6). The total area of soybean growth and production reached nearly 9 × 106 ha and more than 16 × 106 t, respectively, based on data from China’s Bureau of Statistics. As a consequence, fluctuations in soybean growth area and production in China have a large impact on the world supply and price of edible oil and soybeans (Yang et al. 2010). Although soybeans are extensively cultivated across China (Yang et al. 2006, 12–14), approximately 50% of the total area of soybean cultivation in China is located in the northeastern part because of its high productivity (Zhao et al. 2003; Wei et al. 2007). For this reason, we selected northeast China as the study area in this paper.
In the past 30 yr, data from China’s Bureau of Statistics have indicated that the total area and production of spring soybean in northeast China has been steadily increasing. In 1980, the total planted area and production amounts were 265.94 × 104 ha and 334.49 × 104 t, respectively. By 2009, those numbers reached 460.93 × 104 ha and 703.89 × 104 t. Farmers appeared to have expanded the planting area of spring soybeans into highly unsuitable regions because of the potential economic benefit. As a result, in these marginal areas, very low yields per unit area have been achieved, leading to a considerable negative impact on farming communities (Shi et al. 2008). Little attention, however, has been paid to this problem in the literature. Hence, there is an urgent need for the development of suitability mapping for spring soybean cultivation, including a comparative analysis of suitability mapping with actual planted areas.
Assessment of land-use suitability is one of the most important branches in land science. The literature on land suitability analysis is extensive, and, in general, most of the studies focus on biophysical characteristics such as climate, terrain, and soil properties. Allen et al. (1995), for instance, used topography, soil, land cover, and the interrelationships among landform, soil, and vegetation to assess land-use suitability. Steiner et al. (2000) presented a framework for land suitability analysis based on a thorough ecological inventory that allows identifying the constraints and opportunities of land conservation and development. Land resource management systems and visualization techniques in geographic information systems (GIS) have also been widely applied in the evaluation of land suitability (Becker and Bugmann 2001; Kalogirou 2002; Wu et al. 2004; Liu et al. 2005). Nevertheless, these studies were limited because their assessments were implemented to identify whether the land was suitable for a particular application, for instance, a given crop growth.
To obtain such knowledge, some researchers have made efforts to evaluate agrosuitability for specific crops (Pan et al. 1982; Xu et al. 2000; Yan et al. 2008; Qiu et al. 2009). Cutforth and Shaykewich (1990) developed a temperature response function for corn and evaluated corn suitability based on temperature. Zhao et al. (2003) integrated the three climatic factors of temperature, precipitation, and sunshine to analyze winter wheat agroclimatic suitability in Henan Province. He et al. (2011) classified soybean suitability in northern China into three types—early maturity, midmaturity, and late maturity—and applied an integrated method to analyze soybean agroclimatic suitability. The Food and Agriculture Organization of the United Nations recently analyzed the suitability of different kinds of rain-fed and irrigated crops using the agro-ecological zoning (AEZ) methodology (Fischer et al. 2002). The AEZ methodology provides a standardized framework for the characterization of climate, soil, and terrain conditions relevant to agricultural production.
Given that agricultural land suitability depends not only on biophysical characteristics that determine the long-term potential of agricultural production, but also on socioeconomic factors that affect the extent to which such potential is realized, a comprehensive analysis of crop suitability systems becomes necessary. In this study, we seek to achieve three objectives: 1) to develop an integrated indicator system for spring soybean cultivation that accounts for key biophysical and socioeconomic factors, including the effects of climatic factors in the individual development stages of the spring soybean life cycle; 2) to map spring soybean cultivation suitability in northeast China; and 3) to spatially compare a mapped suitability assessment with actual areas under cultivation in order to help guide crop cultivation structural adjustment and improved agricultural sustainability in northeast China.
The remainder of this paper is organized as follows. Section 2 describes materials and methods. Sections 3 and 4 discuss the results. The final section concludes this paper.
2. Materials and method
a. Study area
Our study covers northeast China, including three provinces—Heilongjiang, Jilin, and Liaoning (see Fig. 1)—that are home to more than 100 million people. The region has a total area of nearly 790 000 km2, extending from latitude 38°43′N to 53°33′N and from longitude 118°53′E to 135°05′E. The climate is typically warm and features a semihumid continental monsoon. The area’s topography is characterized by plains, mountains, and hills.
Map of the study area.
Citation: Journal of Applied Meteorology and Climatology 52, 4; 10.1175/JAMC-D-11-0259.1
A single-crop system of spring soybean prevails across the study area (Li et al. 1994). The life cycle of spring soybeans is composed of the five important developmental phases: seed emergence, seedling growth, flower bud differentiation, flowering and pods formation, and pod filling and maturity (Hu et al. 2002; Yang and Yang 2009). Although there are many spring soybean varieties grown in northeast China, their similarity in physiological and phenological characteristics allows us to easily classify soybean as early maturity, midmaturity, and late maturity (Ding 1959; Li and Chang 1998). This classification facilitates the study of agroclimatic suitability of spring soybean (Qiu and Wang 2007, 3–6).
b. Method description
1) Establishment of an integrated indicators system
To evaluate suitability for spring soybean cultivation in northeast China in a holistic way, we combined both biophysical and socioeconomic factors to establish an integrated indicator system (see Table 1). Biophysical factors include four key aspects—climate, terrain, soil, and natural disaster rate. The climate indicator is composed of temperature, precipitation, and sunshine and is further discussed in section 2b(2). The terrain indicator includes slope aspect and gradient. The soil indicator was developed based on soil pH, organic matter content, available phosphorus, available potassium, alkaline hydrolytic nitrogen, and topsoil salinity. The hazard rate is the frequency of occurrence of natural disasters during a given time period, which would affect the spatial distribution of cultivation. Socioeconomic factors include infrastructure, technological improvements, and the farmers’ willingness to cultivate. We chose irrigation accessibility and rural electricity consumption as indicators for infrastructure. Rural electricity consumption not only reflects available electricity in a given field, but the irrigation capacity for growing soybean, as soybean is generally irrigated by electrical equipment in the study area. The importance of technical progress has been widely recognized in agricultural production to facilitate crop cultivation and improve land quality. Here, we employ the use of mulching film and chemical fertilizers, as well as increased mechanization, as proxies for the effects of technological improvements on cropland cultivation suitability.
The integrated indicator system for spring soybean cultivation suitability in northeast China.
Farmers are the decision makers when it comes to crop cultivation. It has been demonstrated that a farmer’s willingness to cultivate depends on the distance from home to fields (He et al. 2010). The idea is that if the location of a given field were to be farther away from a farmer’s residence than are other fields, the farmer would be less likely to cultivate that field. Following He et al. (2010), we adopt cultivation radius as an indicator of the farmers’ willingness to cultivate. The complete indicator system is shown in Table 1.
2) Climatic suitability analysis
Crop growth and yield are largely determined by weather during the growing season; even minor deviations from the usual weather pattern in any growing phase will lead to a change in the efficiency of externally applied inputs and thus have an impact on yield (Mall et al. 2004; Bhatia et al. 2008). Agroclimatic suitability for the life cycle of spring soybeans is mainly controlled by the climatic factors of temperature, precipitation, and sunshine in the different developmental phases of the crop. To calculate agroclimatic suitability for spring soybean cultivation in northeast China, we simultaneously applied crop-response functions for temperature, precipitation, and sunshine on the basis of different phases of the life cycle of spring soybeans. Previous studies suggested that a midmaturity variety is the most suitable choice for northeast China. Thus, the climatic data were processed using developmental dates for midmaturity spring soybeans (He et al. 2011). Since climatic suitabilities obtained from crop-response functions in different phases make different contributions to the overall cropping suitability, the contributions of each phase need to be weighted (Collins et al. 2001; Wei et al. 2007). The weighting method is referenced in the related research reported upon by He et al. (2011). After weighting climatic factors in the different developmental phases, a climatic suitability estimate for the entire life cycle of spring soybeans can be determined.
3) Weighting index
It is clear that each indicator in the system makes different contributions to the overall cultivation suitability for spring soybeans. To account for the relative contribution of the various indicators, each was assigned weighting values. Weighting values were derived using a heuristic approach based on the assessments of 23 experts in related research fields such as land use and cover change, soil science, agronomy, agricultural economy, geography, ecology, and environmental science. These experts were invited to weight the indicators in the study area by means of a questionnaire. The weighting values ranged from 0 to 30. In terms of the Delphi method, if the frequency distribution of the weights from experts conformed to a normal distribution, the weighting could be used. When that was not the case, the 25% of the weights that were most extreme were eliminated and the results were returned to experts for reweighting until conformance to a normal distribution was achieved. After three such rounds, weightings that were considered to be satisfactory were obtained. Afterward, with the application of the weighted statistical method, the final weighting values were obtained (Shi et al. 2008).
Indicators of sunshine, precipitation, and ≥10°C temperature in the previously mentioned reference were too general and were only used for the whole period of crop growth. Thus, in this paper, the three indicators of sunshine, precipitation, and ≥10°C temperature were integrated into one called climatic suitability. The implication of climatic suitability was previously discussed in section 2b(2). The weighting values of the three indicators were added and found to be 0.173, which was assigned to be the new climatic suitability indicator. Expert opinion and weighting values reported in He et al. (2011) were combined to produce the final indicator weighting system presented in Table 1.
c. Data collection and processing
To maintain comparability with previous studies (Shi et al. 2008; He et al. 2009, 2010, 2011), 1 km × 1 km was selected as the spatial unit for this study.
1) Climatic data collection and processing
All the meteorological data used in this study were acquired from 83 China Meteorological Administration standard weather stations located in the study area (see Fig. 2). The meteorological data were the average daily values for the period 1980–2010. Phenological phases were matched by the average recorded values of the three climatic factors during the life cycle of the soybeans. Weather data were interpolated across 1 km × 1 km grids using a kriging function in the GIS software, and agroclimatic suitability in different soybean life cycle phases was calculated (He et al. 2011). Subsequently, the agroclimatic suitability for the whole cycle of spring soybean growth was calculated by weighting agroclimatic suitabilities in the different developmental phases.
Locations of weather stations in the study area.
Citation: Journal of Applied Meteorology and Climatology 52, 4; 10.1175/JAMC-D-11-0259.1
2) Soil data collection and processing
Soil data were collected from the Natural Resources Database for northeast China, which was established when authors implemented a related study described by He et al. (2009). In conducting the above-mentioned project, 742 points in the 0–20-cm surface topsoil at a 10 km × 10 km spatial resolution were surveyed. Attributes such as soil and crop types, soil nutrition, cropping system, topography and geomorphology, irrigation conditions, and facilities to be used for growing soybeans were recorded. The Agro Services International soil nutrients system evaluation method was adopted to test the soil samples and to acquire the chemical test data related to soil nutrition (Hu et al. 2007).
We applied the kriging and cokriging functions within the GIS software to generate 1 km × 1 km spatial distribution maps of pH, organic matter, available phosphorus, available potassium, alkaline hydrolytic nitrogen, and topsoil salinity (Shi et al. 2007).
3) Administrative data collection and processing
The data for hazard rate, irrigation accessibility, rural electricity consumption, mulching film use, and chemical fertilizer use, as well as mechanization extent, were available at the county administrative unit level from the Natural Resources Database for northeast China. We applied GIS software to transfer the attribute data into 1 km × 1 km gridded spatial data (Shi et al. 2008).
Digital elevation model data at a scale of 1:250 000 were obtained from the Institute of Agriculture Resources and Regional Planning, Chinese Academy of Agricultural Science. Slope gradient and aspect and arable land cultivation radius data were also processed onto 1 km × 1 km grids (He et al. 2010).
All indicator values in the 1 km × 1 km data layers were standardized into the range 0–1 (Shi et al. 2008). After weighting the layers, the output of final suitability for cultivating spring soybeans was produced. We applied natural breaks to classify suitability into three types: suitable, moderately suitable, and unsuitable. A comparative analysis was then completed.
To make a comparative analysis of the results for the biophysical and socioeconomic indicators, researchers applied a two-way function using Multi-Criteria Analysis Shell for Spatial Decision Support (MCAS-S), version 3.0, software. Two-way comparison enables the user to create a two-way comparison map, explore the association among input classes, and define a color ramp and value scale to highlight the association between high or low values, or feature a particular geographical region (Hill et al. 2006). At present, the software is open to the public and can be freely downloaded online (http://www.brs.gov.au/mcass).
3. Results
a. Suitability for cultivating spring soybean
A map showing the suitability for cultivating spring soybeans in the northeast China study area is shown in Fig. 3, where the index is classified into suitable (suitability ≥ 0.41), moderately suitable (0.36 ≤ suitability < 0.41), and unsuitable (suitability ≤ 0.36). The average suitability index for cultivating spring soybeans in northeast China was 0.389, implying that the whole study area is positioned in the middle level of the suitability scale for spring soybean growth. Referring to results shown in Fig. 3, it is clear that the spatial distribution of suitability does not match the administrative boundaries, meaning that the results are not constrained by data conforming to administrative boundaries.
Suitability map for cultivating spring soybeans.
Citation: Journal of Applied Meteorology and Climatology 52, 4; 10.1175/JAMC-D-11-0259.1
Areas classified as most suitable for cultivating spring soybeans are located in the Songnen Plain, eastern and central Jilin Province and the Liaohe River Plain, and eastern and central Liaoning Province. These suitable regions occupy approximately 9.09 × 104 km2, accounting for nearly 11.50% of the total area of northeast China. Among areas classified as suitable, 3.32 × 104 and 3.77 × 104 km2 are distributed across Jilin and Liaoning Provinces, respectively, accounting for 36.5% and 41.5% of the total suitable area. Heilongjiang Province only has approximately 22% of the suitable area.
A total area of 7.99 × 104 km2 is classified as unsuitable for spring soybean cultivation, accounting for 10.11% of the total territory of the study area. In contrast to those areas classified as suitable, the unsuitable areas are mainly located in Heilongjiang Province (see Fig. 3). In this province, 6.47 × 104 km2 is classified as unsuitable, accounting for 81% of the total area of the study region classified as unsuitable. Unsuitable areas are located in the Sanjiang Plain, the easternmost parts of Heilongjiang Province, and the fringe areas between the Daxingan Mountain Range, the Xiaoxingan Mountain Range, and the Songneng Plain. Areas of 0.54 × 104 and 0.98 × 104 km2 are classified as unsuitable for spring soybean cultivation in Jilin and Liaoning Provinces, respectively, accounting for 6.76% and 12.24% of the total area of each province. Areas of these provinces classified as unsuitable areas are mainly located in the Changbai Mountain Range.
A total area of 11.45 × 104 km2 is classified as moderate-level suitability, accounting for 14.49% of the total study area. Moderate-level areas for cultivating soybeans are mainly located in Heilongjiang Province with an area of 8.48 × 104 km2, accounting for 74.08% of the province. Jilin and Liaoning Provinces have percentages of 12.55% and 13.37% of the total, respectively. The spatial distribution of areas classified as moderate-level suitability is consistently associated with water flows and rivers.
b. Comparative analysis of biophysical and socioeconomic suitability
The comparative result is shown in Fig. 4. Within the matrix of Fig. 4, the results are categorized into three classes, which are represented by red, green, and blue. The red color indicates that biophysical suitability was better than socioeconomic suitability, and, on the contrary, the green color shows that socioeconomic suitability is better than biophysical suitability. The blue color indicated that the importance of biophysical suitability and socioeconomic suitability are the same. In terms of statistical analysis (Table 2), the red region occupied 44.27% of the total study area (excluding the no-data region), the green region occupied 0.33%, and the blue area was 55.40%. In general in northeast China, biophysical suitability and socioeconomic suitability matched very well, which meant that the two suitabilities in most areas had matching high values or low values. However, in the fringe areas between mountainous areas and the Sanjiang Plain, the biophysical suitability was much better than the socioeconomic suitability.
Comparison between the biophysical suitability and the socioeconomic suitability performed using MCAS-S. Red indicates that the biophysical suitability was better than the socioeconomic suitability, and green shows the opposite.
Citation: Journal of Applied Meteorology and Climatology 52, 4; 10.1175/JAMC-D-11-0259.1
Statistics from a comparative analysis between the biophysical suitability and the socioeconomic suitability in the indicated locations. The numbers in parentheses indicate the class values assigned in MCAS-S (see Fig. 4) and used in the reported statistics.
We further performed a statistical analysis at the provincial administrative level. The results are also reported in Table 2. For Heilongjiang Province the red areas occupied 80.87% of the total red area in northeast China and 60.29% of the study area in Heilongjiang (excluding the no-data region). The red areas were predominantly spatially distributed in the east part; these are the fringe areas between Xiaoxinganling Mountain and Haerbin Plain. Meanwhile, there were some other regions of red scattered in the Sanjiang Plain. Clearly, there is a very large scope to improve agricultural infrastructure or take adaptive technological outputs to increase the productivity of arable land. Despite being one of the most grain-productive provinces in China, the inputs of resources to agricultural production in Heilongjiang are still low. The acreage of paddy rice in Heilongjiang has been decreasing for several years because of a shortage of water irrigation facilities. Considering that Heilongjiang has an excellent endowment of natural resources, the local governments and farmers could make arable land more suitable for cultivation through increased agricultural investment in the form of irrigation, road construction, and so on.
The suitability situation in Jilin Province tells a different story than in Heilongjiang. In Table 2, the red regions only occupied 5.31% of the total red area in northeast China, which is spatially distributed in the central parts of the province. However, the blue areas occupied 29.13% of the total blue area in northeast China and 86.88% of the study area in Jilin (excluding the no-data region). In comparison with Heilongjiang, Jilin has only limited potential for improving its suitability through increasing agricultural inputs. This may be because the government and farmers in Jilin have paid much more attention to agricultural input application and infrastructure construction in the past decades, which also helps to explain why there were always fruitful harvests in Jilin Province, especially for maize. The agricultural infrastructure in Jilin is better than that in Heilongjiang.
For Liaoning Province (see Table 2), the blue regions occupied 28.67% of the total blue area in northeast China and 72.06% of the study area in Liaoning (excluding the no-data region). Meanwhile, the red regions occupied 12.67% of Jilin and 27.75% of the study area in Liaoning (excluding the no-data region). The situation of Liaoning is somewhere between Jilin and Heilongjiang. The potential areas for increasing suitability through increasing agricultural inputs are mainly located in the boundary agriculture–pasture areas between Liaoning and Inner Mongolia Province.
c. Comparative analysis for actual growth and suitable areas
To spatially compare areas classified as suitable for soybean cultivation with areas that are under cultivation, we utilized a grid-based map of soybean cultivation in northeast China, acquired and interpreted from Moderate Resolution Imaging Spectroradiometer (MODIS) imagery by Huang et al. (2010). The original spatial resolution of the imagery was 0.25 km × 0.25 km. After resampling to 1 km × 1 km data layers, we overlaid this dataset with the suitability mapping to detect spatial differentiations. The results, as shown in Fig. 5, suggest that the total area under suitable soybean cultivations in the study regions is 1.14 × 104 km2. Most of these matching areas are located in the Songneng Plain. Of these matching areas, 0.28 × 104 km2 are located in Jilin, 0.19 × 104 km2 are in Liaoning, and 0.67 × 104 km2 are in Heilongjiang. The total area under moderately suitable soybean cultivation is 2.13 × 104 km2. The matching areas are mainly located in the fringe regions between mountains and plains, and the Sanjiang Plain in Heilongjiang. Of these areas, 813 km2 are located in Jilin, 873 km2 are in Liaoning, and 1.96 × 104 km2 are in Heilongjiang. The total area under unsuitable soybean cultivation is 1.33 × 104 km2. These areas are mainly located in the mountainous regions of Jilin and Liaoning and the fringe areas between mountains and plains, as well as the Sanjiang Plain in Heilongjiang. Of these matching areas, approximately 500 km2 are located in Jilin, 289 km2 are in Liaoning, and 1.25 × 104 km2 are in Heilongjiang.
Comparison between the suitability and “actual” growth maps (from satellite data) for spring soybeans in northeast China. The colors define the areas where suitability agrees with “actual” growth.
Citation: Journal of Applied Meteorology and Climatology 52, 4; 10.1175/JAMC-D-11-0259.1
To summarize, of the total area under soybean cultivation in northeast China, 24.78%, 46.30%, and 28.92% are classified, respectively, as suitable, moderately suitable, and unsuitable for spring soybean cultivation, equivalent to 62.21%, 28.41%, and 9.38% for Liaoning, 73.31%, 16.53%, and 10.16% for Jilin, and 15.39%, 51.70%, and 32.91% for Heilongjiang. Relative to Jilin and Liaoning, the prospects for spring soybean cultivation in Heilongjiang are found to be significantly less suitable.
4. Discussion
As has been widely recognized, it is difficult if not impossible to establish and apply a single set of indicators to evaluate crop cultivation suitability at different spatial scales from global level to farm level, due largely to the lack of available data and issues concerning spatial scaling. Different factors have different influences on suitability outcomes at different spatial scales. When implementing a crop suitability assessment for a specific region, the indicator system needs to accommodate not only generally recognized factors such as climate, soil, and topography, but also specific indicators that target factors impacting the practical aspects of cultivation. In this paper, we considered both biophysical and socioeconomic influential factors in establishing an indicator-based approach to evaluating the cultivation suitability for spring soybeans. The method not only involves farmers’ willingness to plant crops and the natural hazard rate on the suitability for cultivating spring soybeans, but it also embraces the effects of temperature, precipitation, and sunshine in different developmental phases of this crop. The output of the evaluation can therefore be used to help guide the implementation of structural adjustments in agriculture in northeast China.
In the study, we have used several types of statistical data organized at the county administrative unit level, such as hazard rate, irrigation accessibility, rural electricity consumption, mulching film use, chemical fertilizer use, and mechanization extent. Although finer spatial resolution of these inputs is desirable, the analysis has demonstrated that the suitability mapping for spring soybean cultivation is not unduly affected or constrained by input information aggregated at the county level. Improved spatialization of socioeconomic data is nevertheless desirable and will be a focus of future work in developing cultivation suitability mapping for crops. At present, a natural break method has been widely used in the research of classifying suitability. In general, the natural break method is an applicable and useful approach for classifying land suitability. When classifying land suitability, there are several aspects that are key; these include the area spatial scale and the methodology and criteria (the same natural breaking points if the natural break method is used) used. When comparing land suitability, the same methodology and criteria at the same scale should be used. In this study, the area of interest was limited to northeast China at the regional scale. Comparisons of results at different spatial scales and classification methods are needed for further studies.
The analysis presented in Fig. 3 clearly leads to the conclusion that areas most suitable for cultivation are generally situated on the plains—for instance, the Songnen Plain—whereas unsuitable areas are generally located in high-elevation regions—for instance, the mountainous areas in Jilin Province. However, an interesting observation is that the Sanjiang Plain, located in the easternmost part of Heilongjiang Province, is classified as unsuitable in relation to agroclimatic suitability. It is well known that soybean growth is sensitive to day length. Insufficient sunshine in three of the five phenological phases—namely the first (emergence), second (seedling growth), and fifth (pod filling and maturity)—greatly affect crop yield. In addition, the combined factors of comparatively low irrigation accessibility, long distance for cultivation, and comparatively high topsoil salinity also lowered the suitability for cultivating spring soybeans in the Sanjiang Plain.
In Fig. 4, in northeast China, biophysical suitability and socioeconomic suitability matched very well. However, in the fringe areas between mountainous zones and the Sanjiang Plain, the biophysical suitability was much better than the socioeconomic suitability. As a result, there is a very large scope for the study area to improve the agricultural infrastructure or take adaptive technological outputs to increase the productivity of soybean cultivation. As a major province for crop cultivation, Heilongjiang has significant potential for improved soybean production, providing more importance is given to agricultural input and infrastructure construction in the future decades.
Results presented in Fig. 5 indicate that spring soybeans are being cultivated in locations that are not suitable for their growth. This helps to explain why spring soybean cultivation has diminished in recent years.
In the past three decades, since the adoption of Chinese reform and “opening up” policies, the acreage dedicated to spring soybean cultivation has drastically increased, driven mainly by large increases in the selling price for soybeans in northeast China. In the 1980s, the average annual cultivated acreage for spring soybeans was 297.33 × 104 ha, but in the recent 10 yr from 2000 to 2009 the area under cultivation soared to 420.17 × 104 ha, a total increase of 41.31%. During this period the price of soybeans dramatically increased from approximately 1 renminbi (RMB) kg−1 to 2.5 RMB kg−1 and, in some instances, more than 3 RMB kg−1. The comparative advantage of growing soybeans rather than other crops has provided a strong incentive for farmers to increasingly reclaim the land for soybean cultivation no matter whether that land was a suitable match for the crop or not. From the standpoint of economic policy, local farming communities should be provided with relief from poverty incurred by the cultivation of spring soybean as soon as possible. To assist in this, we suggest that the government subsidize crop cultivation to an extent sufficient to enable farmers to recoup the losses that result from giving up cultivating soybeans in unsuitable areas. In this way, together with the guarantee of the income to local people, agricultural cultivation structural adjustments could be feasible in practice. The relationship between cultivation suitability and crop growth is complicated; however, agricultural structural adjustments that scientifically target the matching of crop type to cultivation suitability is the goal we are seeking to achieve.
Originally, we envisaged establishing an indicator system to obtain the growth suitability for spring soybeans and to spatialize it. Then the actual spatial distribution of spring soybeans derived from remote sensing technology and GIS was used to validate it. Although this was a complete research process, we found that the actual situation is more complicated than we had imagined. We did some field surveys and talked to farmers in the study area. Through these discussions we learned that in some regions suitable for growing spring soybeans, especially in Heilongjiang Province, rice and maize were grown instead. The reason behind this result is the comparative benefits among the different crops. In other words, the spatial distribution of spring soybeans is determined by not only crop-growth suitability but also the comparative benefits of different crops. So, it is very difficult to make accurate cropping estimates. As our next step, we are planning to work out the growth suitability for maize and paddy rice by using the same methodology and criteria, to then compare the suitability of different crops, and to extract the weighting values of comparative benefits involved with growing the different crops. This should help us make more accurate cropping estimates.
5. Conclusions
In this paper, we describe an integrated indicator-based system used to assess the suitability of cultivating spring soybeans in northeast China using both biophysical and socioeconomic factors. We also compare the assessed suitability for soybean cultivation between the biophysical suitability and the socioeconomic suitability. The findings of this study suggest that the pattern of cultivation for spring soybeans in Heilongjiang is unreasonable and that the Songnen Plain in northeast China is the most suitable for spring soybean growth while the Sanjiang Plain is the least suitable. Our comparative analysis indicates that the biophysical suitability and the socioeconomic suitability matched in the majority of the areas under study. In contrast, in the fringe areas between mountainous regions and the Sanjiang Plain, the biophysical suitability was much better than the socioeconomic suitability; implying that structural adjustment of crop cultivation in the agricultural sector in this region needs to be guaranteed in order to support farming communities.
Acknowledgments
This research was funded by the National Natural Science Foundation of China (41001049), the National Program on Key Basic Research Project (973 Program, 2010CB951501-2), the Key Program of National Natural Science Foundation of China (40930101), the National Basic Research Priorities Program of China (2007FY120100), the China–Australia International Cooperation Project (ABN 24113085695), and the Commonweal Foundation of China’s National Academy (2011142-2).
REFERENCES
Allen, R. B., A. E. Hewitt, and T. R. Partridge, 1995: Predicting land use suitability from vegetation and landform in depleted semiarid grassland, New Zealand. Landscape Urban Plann., 32, 31–42.
Becker, A., and H. Bugmann, Eds., 2001: Global change and mountain regions: The Mountain Research Initiative. IGBP Rep. 49, Stockholm, Sweden, 86 pp.
Bhatia, V. S., S. Piara, S. P. Wani, G. S. Chauhan, A. V. R. Kesava, A. K. Mishra, and K. Srinivas, 2008: Analysis of potential yields and yield gaps of rainfed soybean in India using CROPGRO-Soybean model. Agric. For. Meteor., 148, 1252–1265.
Collins, M. G., F. R. Steiner, and M. J. Rushman, 2001: Land-use suitability analysis in the United States: Historical development and promising technological achievements. Environ. Manage., 28, 611–621.
Cutforth, H. W., and C. F. Shaykewich, 1990: A temperature response functions for corn development. Agric. For. Meteor., 50, 159–171.
Ding, Z. L., 1959: Studies on the bioclimatic adaptation of soybean in China (in Chinese). Acta Agric. Zhejiangensis, 4, 39–53.
Fischer, G., H. T. van Velthuizen, M. Shah, and F. Nachtergaele, 2002: Local agro-ecological assessment for agriculture in the 21st Century: Methodology and results. International Institute for Applied Systems Analysis Res. Rep. 02-02, 119 pp.
He, Y. B., Y. Q. Chen, Y. M. Yao, X. G. Xu, P. Q. Tang, and Q. Y. Yu, 2009: Study on spatial characteristics of the relation between arable land non-agriculturalization and food production—A case study of the provinces in northeast China (in Chinese). J. Nat. Res., 24, 439–447.
He, Y. B., H. J. Tang, P. Yang, Y. M. Yao, Y. Q. Chen, and Z. X. Chen, 2010: An evaluation on land suitability for cultivation in the desertification region under the different policy scenarios (in Chinese). Trans. CSAE, 26 (10), 319–324.
He, Y. B., and Coauthors, 2011: Using an integrated response-function method to explore agro-climatic suitability for spring soybean growth in north China. J. Appl. Meteor. Climatol., 50, 1354–1361.
Hill, M. J., R. Lesslie, R. Donohue, P. Houlder, J. Holloway, and J. Smith, 2006: Assessment of tensions between different claims on the natural resources base in the Australian Rangelands using multi-criteria decision analysis. Environ. Manage., 37, 712–731.
Hu, D. C., X. S. Li, and P. Yang, 2007: Compared study on the correlation of ASI method and conventional analysis method in the determination of available soil P, K, Cu and Zn in the soil in Xuancheng (in Chinese). J. Anhui Agric. Sci., 35, 477–478.
Hu, G. H., J. X. Zhang, and C. Q. Tang, 2002: Growth changing and dry matter accumulation and distribution of spring soybean in north China (in Chinese). Xinjiang Agric. Sci., 39, 264–267.
Huang, Q., H. J. Tang, Q. B. Zhou, W. B. Wu, L. M. Wang, and L. Zhang, 2010: Remote sensing based monitoring of planting structure and growth condition of major crops in northeast China (in Chinese). Trans. CSAE, 26 (9), 218–224.
Kalogirou, S., 2002: Expert systems and GIS: An application of land suitability evaluation computers. Environ. Urban Syst., 26, 89–112.
Kou, J. P., and K. S. Feng, 2008: High-Quality Special Soybean Variety and High-productive Cultivar Technology (in Chinese). China Agriculture Press, 299 pp.
Li, N., M. Gao, Y. H. Fu, and G. Q. Ma, 1994: Analysis on the characteristics of spring soybean in northeast China (in Chinese). J. Jilin Agric. Univ., 16, 86–88.
Li, X. H., and R. Z. Chang, 1998: Cluster and principal component analysis of the spring soybean varieties in China. Acta Agron. Sin., 24, 325–332.
Liu, Y. S., Y. C. Hu, and L. Y. Peng, 2005: Accurate quantification of grassland covers density in an alpine meadow soil based on remote sensing and GPS. Pedosphere, 15, 778–783.
Mall, R. K., M. Lal, V. S. Bhatia, L. S. Rathore, and S. Ranjeet, 2004: Mitigating climate change impact on soybean productivity in India: A simulation study. Agric. For. Meteor., 112, 113–125.
Pan, T. F., D. R. Zhang, W. G. Zhang, and Z. R. Li, 1982: Study on climatic biology of soybean in northeast China (in Chinese). Soybean Sci., 1, 17–28.
Qiu, L. J., and S. M. Wang, 2007: Variety in China (in Chinese). China Agriculture Science and Technology Press, 525 pp.
Qiu, Q., X. Y. Yan, W. Zhang, M. H. Zhang, X. J. Yan, S. Q. Gao, and L. C. Dong, 2009: Studies on adaptability of spring soybean varieties in middle-late-maturity soybean area of north China (in Chinese). Jilin Agric. Sci., 34, 4–6.
Shi, S. Q., Y. Q. Chen, Y. M. Yao, Z. B. Li, and Y. B. He, 2007: Assessing the impact of cultivated land change on grain productive capacity: A case study in the central and western regions of Jilin Province (in Chinese). Res. Sci., 29 (5), 143–149.
Shi, S. Q., Y. Q. Chen, Y. M. Yao, Z. B. Li, and Y. B. He, 2008: Correlation analysis between regional cultivated land change and grain production capacity: A case study in northeast China (in Chinese). J. Nat. Res., 23 (3), 1–8.
Steiner, F., L. McSherry, and J. Cohen, 2000: Land suitability analysis for the upper Gila River watershed. Landscape Urban Plann., 50, 199–214.
Wei, R. J., W. Z. Zhang, X. Y. Kang, and Z. Q. Dong, 2007: Application and establishment of climatic suitability dynamic model of crop in Hebei Province. Agric. Res., 25, 5–15.
Wu, L. X., B. Sun, S. L. Zhou, S. E. Huang, and Q. G. Zhao, 2004: A new fusion technique of remote sensing images for land use/cover. Pedosphere, 14, 187–194.
Xu, X. X., P. Gao, and D. S. Jiang, 2000: Fuzzy analysis on the suitability of precipitation to the agro crop growing in Yanan City (in Chinese). Res. Soil Water Conserv., 7, 73–76.
Yan, Q., J. H. Xu, and G. D. Chen, 2008: A study on the spatial relationship between agricultural land use intensity and agro-climatic suitability in China (in Chinese). Ecol. Sci., 27, 107–113.
Yang, L., M. Zhang, and L. B. Wan, 2006: High Quality and Efficient Cultivar (in Chinese). Shandong Science and Technology Press, 106 pp.
Yang, X. F., and D. G. Yang, 2009: Preliminary studies on establishing climatic adaptability index system for spring soybean in northeast China (in Chinese). Seed World, 11, 36–38.
Yang, X. F., D. G. Yang, Y. H. Tang, and Z. H. Wang, 2010: Changes of effective accumulated temperature and ecological suitability of soybean in Heilongjiang Province (in Chinese). Crops, 2, 62–65.
Zhao, F., H. S. Qian, and S. X. Jiao, 2003: The climatic suitability model of crop: A case study of winter wheat in Henan Province (in Chinese). Res. Sci., 25, 77–82.
