Research on the Comprehensive Evaluation of Alfalfa Management in Zuli River Basin

： Taken Alfalfa as an example, this study analyzed the impact of returning cultivated land to grassland in Zuli River basin. The entropy weight method was used to appraise the growth years for alfalfa. The results showed that the appropriate years for alfalfa in the northern of the basin is nine and sex for the southern of the basin. After the recommended planting years, the composite index significantly decreased. Then, the yield of alfalfa in Zuli river basin was expressed in map using multiple regression method.


Introduction
Medicago Sativa is a good forage grass widely cultivated in the world, which has the largest planting area in China [1].With the Characteristics of drought tolerance, cold resistance, salt resistance and barren resistance, alfalfa is a kind of ideal coarse fodder for livestock.With strong adaptability, high yield and rich nutrient, alfalfa not only plays an important role in Chinese traditional agriculture and animal husbandry, but also taken as the preferred species for the artificial grassland in the loess plateau [2].Today in the construction of the western environment, alfalfa was used for returning farmland to grassland, optimizing the agricultural structure and so on [3].
People paid more attention to dried soil layer caused by long term cultivate alfalfa, although there are many advantages.The infiltration depth of rainfall is generally 100-300cm on the Loess Plateau area, while the soil evaporation transpiration depth is up to 800-1000cm, which resulting water deficit in the deep soil, and vegetation degradation [4].
In the recent ten years, with the increase of alfalfa planting area, the Chinese scholars have also studied the influence of Alfalfa on soil condition, the influence factors of alfalfa yield, the influence of Alfalfa on soil moisture and so on [5,6,7,8,9,10,11].In this study, the effects of different rainfall and temperature conditions in the northern and southern part of the river basin, the influence of the soil condition, and the relationship between the planting time and the yield of alfalfa were analyzed to study a recommended management plan for Alfalfa in the basin.And the yield distribution of Alfalfa was simulated by GIS.

Study on the suitable planting time of Alfalfa in Zuli River Basin
Data was collected from the literatures [2], [12].The entropy weight method was used.The principle of availability and identity was taken when choose the indicators.Select common indicators between two experimental points from the indicators obtained.The representative of soil condition was also taken into consideration(Table 1，Table 2).

Spatial distribution of Alafalfa Yield
All of the Alfalfa yield data are collected form the literature(Table 3).
The yield data in this study were derived from the experiment without fertilization and irrigation treatment, or control data in experiment.The ratio in the literature [2] was used to calculate the dry weight for only fresh weight.
Statistical analysis: the environmental factors which were significantly correlated with yield were selected to simulate the yield of Alfalfa in the watershed by the multiple regression model.Regression models of alfalfa production and regional environmental variables were created by stepwise regression method.
Based on the regression model, the grid computing module was used to simulate the spatial distribution of alfalfa yield in ArcGIS, then the yield distribution map of alfalfa was obtained.According to the entropy method, the weights (wi) for each index in Zhenyuan experimental station were calculated and showed in Table 4.The sequence according to the weight is as follows: yields(0.8317),available nitrogen(0.1095), soil moisture content (0.0574), soil bulk density(0.0014).On the basis of the principle of entropy method, the result shows that the change of yield was the biggest, the influence on available nitrogen was the second, and the change range of soil bulk density was the least with the increase of planting years of alfalfa.
Then the comprehensive evaluation was applied to evaluate the growing years for alfalfa with the obtained weights.Table 5 is the comprehensive evaluation value of each year in Zhenyuan experimental station, which shows that: 6 years >8 years >12 years >14 years >18 years >26 years.The conclusion shows that the best growth time for alfalfa in this area was 6 years, 4 years is the next, followed by 8 years.After 8 years, the composite index decreased gradually with the increase of alfalfa growth.According to the entropy method, the weights (w i ) for each index in Zhonglianchuan experimental station were also calculated and showed in Table 6.
The sequence according to the weight is as follows: yields(0.7122),soil moisture content (0.155), available nitrogen(0.1301),soil bulk density(0.0027).On the basis of the principle of entropy method, the result shows that the change of yield was the biggest, the influence on soil moisture content was the second, and the change range of soil bulk density was the least with the increase of planting years of alfalfa.
That different from Zhenyuan experimental station where with more precipitation.
This result shows that the effect of alfalfa's growth year on soil moisture in Zhonglianchuan is more than in Zhenyuan.Similarly, the change of soil bulk density was the smallest, showed that the effects of alfalfa's growth year on soil bulk density is the least, or a more long-term process than other factors.
Then the comprehensive evaluation was also applied to evaluate the growing years for alfalfa with the obtained weights.Table 7 is the comprehensive evaluation value of each year in Zhonglianchuan experimental station, which shows that: 9 years >5 years >3 years >15 years >21 years >25 years.The conclusion shows that the best growth time for alfalfa in this area was 9 years, 5 years is the next, followed by 3 years.After 15 years, the composite index decreased gradually with the increase of alfalfa growth.

Spatial distribution of Alafalfa Yield
Through correlation analysis found that the yield was significantly positive correlated with the mean annual rainfall, the mean temperature in July, average temperature (p<0.05)andaverage rainfall in July, rainfall in May to July(p<0.01).
And the correlation with other environmental factors did not reach a significant level(Table 8).
The abovementioned factors were used as independent factors to establish the multiple regression model of alfalfa production by the backward regression method.
Regression models are follows: y=262.709+44.531x6 +0.781x 4 (1)  The image shows that the yield of alfalfa in the south of the basin is the highest, which peaked out at 2699.12kg/hm 2 .The yields were reduce further and have latitude belt distribution characteristics.The yield of alfalfa in the northwest of the basin is the lowest which was 2072.21kg/hm 2 .

Conclusions
The entropy method which is an objective weighting method was adopted in the comprehensive evaluation of alfalfa growth years.Compared with other methods, it is possible to reduce the interference of human subjectivity to the evaluation process and fully tap the objective differences in each index in the evaluation system.Therefore, this method could more objective to reflect the influence of alfalfa growing period on soil and its own output.Judging from the weight of the entropy method, the weight of yield is much higher than the other three indicators both in the northern and southern parts of the basin, which reflects that the change between the yield was the largest in several indicators and played the largest role in the comprehensive evaluation.
The weight of soil bulk density was small in both two experimental stations, which shows the effect of alfalfa growth on soil bulk density was small compared with the other three indicators.This influence was slightly larger in the northern area with lower rainfall and temperatures than in the southern part of the region with relatively high rainfall and temperature.Although this effect is small, it can not ignore the effects of alfalfa growth on soil porosity.
The weight of soil moisture in Zhonglianchuan experimental station was larger than in Zhenyuan experimental station, illustrated that the effect of alfalfa growth on soil water content was higher in the northern area with less rainfall than in the southern part of the region with relatively high rainfall.According to the results of this study, the cropping years for planting alfalfa is suggested to be 9 years in the northern area, and 6 years in the southern area is appropriate, the longest is 8 years.
According to the correlation analysis of alfalfa yield and multiple environmental factors, the regression equation of yield between mean temperature in July and of rainfall in 5, 6, 7 three months was obtained.The yield of alfalfa in the basin is simulated based on the equation and map was expression in ArcGIS.The result reflects yield of alfalfa was between 2072kg/hm 2 and 2677kg/hm 2 under the artificial management, and decreased from south to north.

Table 1 .
The yields of alfalfa, soil moisture content, soil bulk density and potential mineralized nitrogen in Zhenyuan Experimental Station

Table 2 .
The yields of alfalfa, soil moisture content, soil bulk density and potential

Table 3 .
Data source of the yields of alfalfa and the location

Results and Discussion 3.1 Study on the suitable planting time of Alfalfa in Zuli River BasinTable 4 .
The weights for each indicators in Zhenyuan Experiment Station

Table 5 .
The comprehensive ecaluation values for the different growing years' alfalfa

Table 6 .
The weights for each indicators in Zhonglianchuan Experiment Station

Table 7 .
The comprehensive ecaluation values for the different growing years' alfalfa

Table 8 .
x 6 is temperature in July，x 4 is precipitation in May, June and July.The regression model passed through F test (p=0.018),Correlation analyses between Alfalfa yield and meteorological * Indicates significant correlation between Alfalfa yield and Corresponding Environmental factors(p<0.01). *