Effects Of Farmland Landscape Pattern On Soil Organic Carbon In Lower Liaohe Plain

In the process of farmland utilization,different land use types are gradually formed,and then different landscape patterns of farmland are formed.Different landscape patterns of farmland have different effects on material energy circulation in farmland,and the suitable landscape structure can provide more stable space for good material circulation.As an important index of farmland soil fertility evaluation,soil organic carbon in farmland is not only related to crop yield,but also has an impact on global carbon cycle and climate warming.The main factors affecting soil organic carbon have always been the focus of research.In the past,the factors of soil organic carbon mainly focused on natural factors and micro human activities,while the effects of landscape patterns and changes on soil organic carbon were still lacking.The farmland landscape pattern reflects the basic characteristics of farmland land use.Studying its influence on soil organic carbon can reveal the mechanism of human activities on soil organic carbon at the landscape scale.For this reason,based on the soil sampling points in Lower Liaohe Plain in 1980 and 2010,with landscape pattern index as the main variable and terrain,climate,topographic humidity and vegetation elements as auxiliary variables,correlation analysis and multiple collinearity test were used to screen the variables associated with soil organic carbon.In order to provide theoretical basis for land use planning,landscape layout and soil organic carbon protection in the study area,soil organic carbon was predicted on the premise of correlation and influence relationship.The study concluded that:1.The average soil organic carbon in 1980 was 10.93 g/kg,2010 the average soil organic carbon in the year was 10.78 g/kg,and the average soil organic carbon in the farmland of Liaohe Plain decreased by 0.15 g/kg in the last 30 years2.This paper takes 4 km2 as the optimal range to analyze the characteristics of landscape pattern change in each period.Under the 4 km2 range,the farmland landscape in 1980 showed that the patches in the landscape were more aggregated and less discrete;the largest patch area in the landscape was moderate,the patch shape slightly deviated from the square;the cultivated land in the landscape dominance degree is not obvious,the distribution of each patch type in the landscape is more uniform,and the degree of fragmentation is not high.The farmland landscape pattern index of 2010 has changed,the COH?LPI ?LSI?NP?pd index has increased,and the DIV?SHDI?SHEI index has decreased.Compared with 1980,the overall landscape pattern changes in the study area showed that the aggregation degree between patches increased and the dispersion degree decreased;the maximum patch area proportion increased,and the patch shape changed slightly in an irregular direction;the number of farmland patches in the landscape increased,the diversity and evenness of patch type composition decreased,and the dominant patch area increased.3.The correlation between farmland landscape pattern index and soil organic carbon was different in different periods.SHDI 3km?LPI 3km?SHEI 4km?COH 5km of the pre-selection variables in 2010 had a significant correlation with soil organic carbon formation.a spatial local analysis of the associated farmland landscape pattern indices using the geo-regression weighted model shows that SHDI 3km spatial influence regression coefficient is a high value area in the western part of the study area and a positive correlation between the shannon diversity index and farmland soil organic carbon formation in the western part of the study area;LPI 3km spatial regression coefficient distribution is more uniform,the difference is not obvious in different regions;the SHEI 4km spatial influence regression coefficient is positive correlation and large value in northeast and southwest regions of the study area,indicating that the more uniform patch distribution in this area,the more favorable soil organic carbon accumulation;COH 5km spatial influence regression coefficient values decrease step by step from north to south,and the influence regression coefficient between the southern index and soil organic carbon in the study area is negative,indicating that the accumulation between patches in this area is not conducive to soil organic carbon accumulation.4.Modeling and analysis using auxiliary variables related to soil organic carbon in 1980,and the result of model accuracy R2 was 0.299.For modeling variables including landscape pattern index in 2010,the 2010 model R2 was 0.421.With the exception of altitude elements,the two-stage model uses the same auxiliary variables,the difference is that the 2010 model adds SHDI 3km?LPI 3km?SHEI 4km and COH 5km landscape pattern index.The accuracy of the model in 2010 is larger than that in 1980,which indicates that the landscape pattern index can help to improve the accuracy of the model under the premise of correlation.The soil organic carbon in the study area was predicted based on the 2010 model correlation and influence results.The results showed that the maximum value of soil organic carbon in the study area was 26.83 g/kg,the minimum value was 1.01 g/kg,and the distribution of soil organic carbon decreased from southeast to northwest.