Analysis on Temporal and Spatial Variation of Landscape Ecological Security in Modern Agricultural Area
高星,刘玲,吴克宁,于潇,魏洪斌,宋英赫
DOI: https://doi.org/10.11975/j.issn.1002-6819.2016.08.036
2016-01-01
Abstract:Modern agriculture is the foundation of national economic and social development, and because of its close correlations with nature, the landscape ecological security of modern agricultural region plays an important role in China’s ecological reconstruction. The concept of landscape ecological security (LSES) can be described in both special and general ways. Based on the special definition, LSES is evaluated by the level of natural or human disturbance and the vulnerability of each landscape type in typical model. In this model, the level of disturbance is basically measured by landscape pattern indexes such as landscape fragmentation, isolation, dominance and diversity. The vulnerability of each type of landscape is mostly decided by expert experiences. Thus, the disadvantage of typical LSES evaluation model is that it cannot evaluate the ecological effects of the change of landscape types. Based on the general definition of LSES, some studies introduce P-S-R model on LSES evaluation. The advantage of P-S-R model is that it offers a wider range of index options. However, some of these indexes are based on economic and social statistic data, and they are limited by time and spatial scale of statistics. For these reasons, we built a landscape pattern security index (LSPS) based on the traditional LSES model to evaluate the disturbance of nature or human on landscape pattern and the vulnerability of landscape types. Ecological quality index (EQ) was introduced as a supplement to evaluate the effects of the disturbance on the health and service function of ecosystem. With this method, the LSES change of modern agricultural region during last 30 years was analyzed by a case study on the state-owned Youyi farm in the Sanjiang plain, and the four phases of Landsat TM/OLI images in the year of 1984, 1995, 2004 and 2014 were the main data sources. The major land use/cover types of study area were reclassified into seven classes including construction land, dry land and paddy field (arable land), grass land, water area, natural wetland and forest. Results showed that during the last 30 years, the intense arable land exploitation had resulted in an obvious land use and cover change, which had led to a series of positive or negative ecological consequences. The proportion of arable land area had increased from 54.61% to 76.58%, and natural wetland had declined from 22.24% to 1.60%. By an analysis on land use change matrix, the major type of land use/cover change during the whole period was a conversion from the natural wetland to the dry land, and finally to the paddy field. The time variation trend of LSES in the study area was obviously regular, and the value of LSES in the year of 1984, 1995, 2004 and 2014 were 0.65, 0.57, 0.55 and 0.60 respectively. From 1984 to 1995, the regional LSES showed a decline trend. Because of the implementation of land consolidation, which had basically restored the regional landscape fragmentation, the descent rate of LSES from 1995 to 2004 had slowed down. However, natural wetland loss caused by massive cultivated land expansion still led to a decline of LSES. From 2004 to 2014, more than half of dry land had converted into paddy field during the implementation of land consolidation project. Therefore, the increasing coverage of artificial wetland had, to some extent, restored the loss of ecosystem service value caused by the diminishing natural wetland. During the whole study period, the spatial autocorrelations of LSES were significantly (P<0.05) clustered. The value of Moran’s I increased from 0.5548 of 1984 to 0.5851 of 1995, and declined to 0.2564 of 2014. The descent trend of Moran’s I during the lasts 20 years might be caused by the continuous implementation of land consolidation which had changed the original structure of landscape ecosystem, and the unstable of newly formed pattern of LSES. The result of local autocorrelation of LSES showed that the area percentage of H-H and L-L kept declining, and the area percentage of no significance kept increasing during the last 30 years. This might attribute to the land use/cover change caused by modern agriculture development that had brought negative effects on the landscape ecosystem, and land consolidation that also had restored the fragmented landscape pattern and damaged ecosystem service function. Thus, in order to realize the sustainable development of modern agriculture, the future land consolidation should reinforce the multiple objectives of improving the quality of arable land as well as ensuring the regional ecological security. Verified by the empirical study, the method we proposed can be used to accurately analyze the succession trends of landscape ecosystem caused by natural or human disturbance at both overall and local levels, and the method has a broad applicability on landscape ecological security evaluation in other regions.