| The project of return marginal farmland to forest is the vastest ecological construction project of area,scope,investment,policy.It has curbed the deterioration of ecological environment in Loess Plateau and improved fragile and degraded ecosystems.But up to now,vegetation construction in Loess Plateau faces many challenges owing to various factors such as natural and socio-economic conditions.Our study area was located in Wuqi County,a typical water-wind erosion crisscross region on the Loess Plateau.The area of about 60 km2 was taken as the research subject,and the period from 2000 to 2017 was taken as the research period.Based on the field investigation,the regional vegetation restoration status was studied,and redundant analysis was introduced to explore the driving forces of vegetation coverage change.At the same time,four common afforestation species,including Robinia pseudoacacia(RP),Pinus tabulaeformis(PT),Armeniaca sibirica(AS),and Hippophae rhamnoides(HR)were chosen for the study of adaptability.The variation of the distribution range of each tree species after 10 years from 2007 was analyzed.And combined with Random Forest model,the suitable site conditions for each tree species were proposed.This study contributes to the evaluation of the effectiveness of vegetation restoration in the project of return marginal farmland to forest or grassland.And the results can provide theoretical basis for afforestation planning and design,and also give some suggestions on how to implement sustainable management strategy.The main conclusions are as follows:(1)The research area was rapidly advancing urbanization construction.At the same time,it was also actively returning marginal farmland to forest and restoring vegetation.In general,the results of vegetation construction remained good.From the perspective of land use change,the area of grassland and farmland had been significantly reduced,and the land for forestry and construction had been significantly expanded.The comprehensive dynamics of land use were as follows:the first phase from 2000 to 2007(3.92)>the second phase from 2007 to 2017(1.34)>the total from 2000 to 2017(0.82).From the perspective of vegetation cover change,the high vegetation coverage area with NDVI>0.6 in the study area showed an overall growth trend,which was reflected in 2017(50.36%)>2007(24.98%)>2000(16.04%).In the first stage,11.39%of the study area showed a significantly improvement trend(Slope>0 andp<0.01,Slope>0 andp<0.05),and in the second stage,the value was 3.71%.Overall,20.55%of the area showed a significantly improvement trend.(2)From the perspective of topographic distribution characteristics of vegetation restoration,the average NDVI in the range of altitude 1395m-1485m and slope 25°?35° was higher,and the proportion of areas with high vegetation coverage and significantly improved vegetation coverage was also larger.In addition,the average NDVI of the areas with azimuth angles of 0°?22.5°(north slope)and 315°?360°(northwest slope,north slope)were higher.Within the above range,the area of high vegetation coverage in all elevation zones,slope zones,and slopes increased over time.(3)Redundancy analysis analysis showed that the driving factor which had a significant impact on changes in vegetation coverage were elevation(p=0.002,F=21.51),the shortest distance between samples and roads(p=0.018,F=6.66),slope(p=0.022,F=5.848)and topographic roughness(p=0.048,F=4.398).The factors that were positively correlated and negatively correlated with the trend of vegetation change in different periods were different,and the cause of this result was related to the human activities in the study area.(4)The growth areas of the three arbors became larger,and the HR,as a shrub,showed a decreasing trend.Qualitative analysis of the reasons for changes in tree species distribution revealed that the influence of different topographic factors on the AS was more regular.For RP,PT and HR,although the regularity was unclear,many topographic factors in the same degree of change had similar properties.For example,the slopes of the three vegetation with "significant deterioration" were large;the slope and relief of HR and PT were similar in the same degree of variation.The random forest models of the four tree species established in this study had high accuracy.It was found that slope direction,elevation and the shortest distance between samples and roads impacted each species significantly.Specifically,the most influential factor for RP was elevation,the most important factor affecting PT was the shortest distance between samples and roads.The factor that had the greatest impact on AS and HR was aspect.(5)Generally speaking,RP was most suitable for growth in the azimuth angle of 318°?340°(north north slope and north slope),elevation of 1475m?1515m;PT was most suitable for growth in the azimuth angle of 50°?68°(east-northeast slope),elevation of 1 less than 1407m;AS was most suitable for growth in the azimuth angle of 294°?326°(northwest slope),elevation of 1474m?1497m;HR was most suitable for growth in the azimuth angle of 353°-348°(north slope),elevation of 1 1492m?1498m,and 1525m?1545m.Three kinds of arbors basically showed that they were suitable to grow in areas close to the road,especially for PT,the closer the distance to the road,the better the growth.But HR was opposite of arbors.The reason for this may be that the arbor needs artificial protection,and the HR was mostly shrub in the study area,which had strong root abilities and didn't require excessive management.In addition,all tree species were in poor condition on the southeast,south and southwest slopes The slope range suitable for the growth of AS was the widest(205°),and and the slope range suitable for the growth of PT was the narrowest(100°).the HR was 153°,and the RP was 120°.And the suitable altitude range for each species ranged from wide to narrow:PT(198m),HR(1 18m),AS(109m)and RP(108m). |
PDF Full Text Request