【作者】 张华兵；

【导师】 刘红玉；

【作者基本信息】 南京师范大学 ， 环境地理学， 2013， 博士

【摘要】 盐城淤泥质海滨湿地是我国乃至世界上为数不多的典型原始海岸湿地,在全球生物多样性保护中具有重要战略地位。多年来,在自然和人类社会经济发展影响下,区域湿地景观结构与功能发生巨大变化。尤其是沿海开发战略的实施,致使湿地保护与开发之间的矛盾日益加大。在此背景下,如何辨识自然和人为影响下海滨湿地景观演变空间模式,揭示内在生态过程驱动下的湿地景观结构与格局时空演变机制,是实现海滨湿地合理利用与保护需要解决的重要科学问题。本文综合运用景观生态学、地理学、环境科学等学科理论,以及RS、GIS、数学建模等方法,以盐城海滨湿地典型区域为案例区域,并根据区域湿地景观特征,人为将其划分为人工管理区和自然条件区两部分,从景观生态过程角度,对两个区域湿地土壤属性特征及其变化进行深入研究,揭示不同景观类型生态演变的阈值效应。在此基础上,利用湿地景观结构空间数据和景观生态过程数据,构建基于过程的景观演变模拟模型,对两种发展模式下湿地景观时空演变过程与机制进行系统模拟研究,并对未来区域景观演变进行情景分析与预测研究。该研究不仅体现了基于景观过程模型研究方法上的创新,而且对指导区域湿地生态保护与合理利用具有重要意义。得出基本结论如下：(1)海滨湿地景观结构与格局时空变化明显。区域自2000～2011年期间,景观结构变化主要表现为芦苇沼泽和米草沼泽面积不断扩张,碱蓬沼泽面积不断减小的趋势。其中,人工管理区芦苇沼泽和米草沼泽面积分别增加了30.92%和352.61%,而碱蓬沼泽面积减少了78.81%；自然条件区芦苇沼泽和米草沼泽面积分别增加了368.25%和96.63%,而碱蓬沼泽面积减少了60.96%。景观格局变化上,主要表现为芦苇沼泽向海扩张,米草沼泽向海和陆两个方向同时扩张,碱蓬沼泽则以向中心收缩为主。其中,人工管理区芦苇沼泽向海扩张速度为240m/年；米草沼泽向海陆方向扩张速度分别为40m/年和61m/年,以向陆地方向扩张为主;碱蓬沼泽从海陆两个方向向中心收缩的速度分别为61m/年和240m/年,以向海洋方向收缩为主。自然条件区芦苇沼泽向海扩张速度为162m/年；米草沼泽向海陆方向扩张速度分别为117m/年和40m/年,以向海扩张为主；碱蓬沼泽从海陆两个方向向中心收缩的速度分别为30m/年和162m/年,以向海洋方向收缩为主。(2)土壤性状是控制海滨湿地景观类型空间分异的主要要素,其中土壤水分和盐度空间分异是影响不同景观类型演变的主导因子。总体上,从陆向海方向,芦苇沼泽、碱蓬沼泽和米草沼泽的土壤水分和盐度均呈现递增趋势。其中,人工管理区,在干旱年份,芦苇沼泽、碱蓬沼泽和米草沼泽土壤水分平均含量分别为38.83%、41.79%和46.97%,盐度平均含量分别为0.39%、0.71%和1.76%。在湿润年份,各个景观类型水分含量均略有增加,分别增加了0.44%、4.09%和1.51%,而盐度上各种景观类型则呈现减少特征,分别减少了28.21%、36.62%和32.39%。自然条件区,在干旱年份,芦苇沼泽、碱蓬沼泽和米草沼泽土壤水分平均含量分别为36.79%、40.70%和44.16%,盐度平均含量分别为0.43%、0.93%和1.34%；在湿润年份,芦苇沼泽和米草沼泽土壤水分略有增加,分别增加了5.60%和4.23%,碱蓬沼泽的土壤水分略有减少,减少了0.69%,而盐度在各种景观类型上呈现减少的特征,芦苇沼泽、碱蓬沼泽和米草沼泽土壤盐度分别减少了11.63%、32.26%和35.07%。(3)控制海滨湿地景观演变的土壤水分和盐度具有明显的阈值效应。研究表明,不同景观类型水、盐阈值具有明显的差异性。其中,芦苇沼泽水分阈值范围为33.11%-42.28%,盐度阈值范围为0.15%-0.53%；碱蓬沼泽水分阈值范围为33.11%-48.63%,盐度阈值范围为0.53%-0.89%；米草沼泽水分阈值范围为26.42%-55.33%,盐度阈值范围为0.89%～1.44%；光滩水分阈值范围为48.63%-66.59%,盐度阈值范围为0.31%～0.89%。(4)综合利用区域景观结构空间分布数据和土壤性状数据,构建了基于过程的景观演变模拟模型。该模型不仅具有动态显示区域景观时空演变能力,而且能够从景观生态过程变化角度,揭示区域景观演变机制问题。经检验,模型在模拟区域景观演变过程中表现出较高的准确性,总体精度可达85%以上。(5)根据区域发展和景观变化特征,设置三种不同情景,利用景观过程模型对区域湿地景观演变进行模拟和预测研究。其中,现状模式(情景Ⅰ)情景模拟表明,碱蓬沼泽是受影响最为显著的湿地类型。在芦苇沼泽和米草沼泽持续迅速扩张影响下,人工管理区碱蓬沼泽将于2025年基本消失；自然条件区碱蓬沼泽将于2030年也基本消失。生态恢复模式(情景Ⅱ)情景模拟表明,人工管理区在去除堤坝影响下,碱蓬沼泽面积减少变缓,至2020年碱蓬沼泽的面积较情景Ⅰ增加了10.67倍;自然条件区在人工恢复芦苇沼泽情景下,至2020年,芦苇沼泽面积由情景Ⅰ增加了3.53%,碱蓬沼泽面积较情景Ⅰ增加了150.07%；米草沼泽面积较情景Ⅰ减少了7.90%。保护本地物种碱蓬模式(情景Ⅲ)情景模拟表明,至2020年,在控制米草向陆扩张情景下,人工管理区和自然条件区碱蓬沼泽面积分别是情景Ⅰ的3.26倍和5.46倍；在去除互花米草情景下,人工管理区和自然条件区碱蓬沼泽面积分别是情景Ⅰ的7.24倍和20.65倍。更多还原

【Abstract】 The muddy coastal wetland in Yancheng is one of the most typically primitive coastal wetland for the significant strategic position in global biodiversity protection.. Over the years, under the influence of the nature and human socio-economic development, the wetland landscape structure and function changed dramatically. Especially, the contradictions between wetland protection and development gradually increase for the implement of the strategy about the Coastal Development Drive. In this context, how to identify the spatial patterns of the landscape evolution under the influences of natural and human factors and revealing the mechanisms for the wetland landscape structure and the pattern of spatial-temporal evolution under the intrinsic ecological process drive. They are important scientific problems to achieve the aim that using and protecting the coastal wetlands reasonable.The theories about the landscape ecology, geography, and environmental sciences and the methods of RS, GIS, and mathematical modeling are used in this study. Taking the coastal wetland in Yancheng as the typical area, and the region was separated into the labor control and natural condition parts to investigate the soil properties and changes, and to reveal the threshold effects of different ecological evolution from the perspective of the landscape ecological process. Based on these researches, a landscape model based on the ecological process was constructed by using the spatial data of the landscape structure in wetland and the ecological process data of landscape. The spatial-temporal evolution process and mechanisms were studied under the two modes. And the scenario analysis and prediction for the evolution of regional landscape were implemented. The study not only embodies the innovation of the research methods about the landscape process model, but which is of great significance for guiding regional wetland ecological protection and rational utilization. The main conclusions are as follows:(1)The temporal-spatial changes of landscape structure and pattern in the coastal wetlands are obvious. Landscape structural changes appeared the trends that the reed swamp expanded continuously and the area of spartina marsh decreased. The areas of reed marshes and Spartina alterniflora marsh by the manual management increased by30.92% and352.61%, respectively, while that of Suaeda marsh decreased by78.81%; the areas of reed marshes and Spartina marsh area under natural conditions, increased of368.25%and96.63%, respectively. While that of Suaeda marsh decreased by60.96%.On the aspect of landscape pattern, it appears that reed marsh expanded to the sea and Spartina alterniflora marsh expanded to the sea and land in both directions at the same time, and the Suaeda marsh contracted to the center. Among these, in human activities area, the expansion speed of reed marshes is240m/year, that of Spartina alterniflora marsh are40m/year and61m/year to the sea and land, respectively, and the expand to land is the main trend. The contraction speed of the Suaeda marsh from land and sea are61m/year and240m/year, respectively, and expand to sea is the main trend. In natural conditions, the expansion speed of reed marshes is162m/year, and that of Spartina alterniflora marsh to land and sea are117m/year and40m/year, respectively, and the main treed is to sea. The contraction speed of Suaeda marsh from land and sea are30m/year and162m/year, respectively, and the main trend is contracting to the sea.(2) Soil properties are the main elements that controlling the landscape type spatial differentiation and the soil moisture and salinity spatial differentiation are the dominant factor affecting the evolution of landscape types. Overall, the direction from land to sea, the soil moisture and salinityof the reed swamp, Suaeda marsh, and Spartina alterniflora marsh, showed an increasing trend. In the dry years, the average soil moisture content of the reed marshes, Suaeda marsh and Spartina alterniflora marsh are38.83%,41.79%and46.97%in human management areas, respectively. And the salinity average contents in these three landscapes are0.39%,0.71%and1.76%, respectively. In wet years, the moisture content of the corresponding landscape types are slightly increased of0.44%,4.09%and1.51%, respectively, and salinity are decreased by28.21%,36.62%and32.39%.In the natural conditions area, in dry years, the average soil moisture contents of the reed marshes, Suaeda marsh and Spartina alterniflora marsh are36.79%,40.70%and44.16%, respectively, and the average salinity contents are0.43%,0.93%and1.34%, respectively. In the wet year, the soil moistures of the reed marshes and Spartina alterniflora marsh increased by5.60%and4.23%, respectively, and the soil moisture of Suaeda marsh reduced by0.69%, the salinity in all kinds of landscape types decreased, and the decreased number are11.63%,32.26%and35.07%, respectively.(3) The threshold effect is obvious by controlling the soil moisture and salinity of the coastal wetland landscape. The study results show that the range of threshold of soil moisture and salinity are different in different landscape types. The scope of the soil salinity for the reed swamp is0.15%～0.53%, and that of the soil moisture is33.11%～42.28%; the scope of the soil salinity of the salsa swamp is0.53%～0.89%, and that of the soil moisture is33.11%～48.63%; the scope of the soil salinity of the spartina marsh is0.89%-1.44%, and that of the soil moisture is26.42%-55.33%; the scope of the soil salinity of the optical flat is0.31%～0.89%, and that of the soil moisture is48.63%-66.59%.(4) A simulation model of landscape evolution was constructed based on the process through using the spatial distribution data of regional landscape structure and soil properties data. The model not only possesses an ability of visualizing thedynamic changes of the spatial and temporal evolution, but also can reveal the evolution mechanism for the regional landscape from the perspective of the changeed landscape ecological processes. Based on the validation, the model showed a high accuracy in simulating the regional landscape evolution, and the overall accuracy can reach up to85%, and even more.(5) According to the characteristics of regional development and landscape changes, three different scenarios were designed, and the regional landscape evolution was simulated and predicted by using landscape process models in the coastal wetlands. The scenario simulation of the status quo mode (scenario I) shows that the Suaeda swamp is most affected types in the wetlands. Under the trend of the continued rapid expansion of reed marsh and Spartina alterniflora marsh, the Suaeda swamp will almost disappear in2025in the artificial management area. The Suaeda marsh will disappear in2030in the natural conditions area. The scenario simulation of the ecological recovery mode (scenario II) shows that the decrease trend of the Suaeda marsh area will be slow, and the area of the Suaeda marsh will increased by10.67times compared to scenario I in the artificial management area under the removal of the dam’s influence to2020. Under the scenario of the artificial restore area, the area of reed swamp will increase3.53%, and the area of the Suaeda marsh will increase150.07%compared to scenario I, the area of the Spartina marsh will decrease by7.90%compared to scenarios I in the natural condition to2020. The scenario simulation of protecting the local species Suaeda mode (scenario III) shows that controlling the expansion of the Spartina landward to2020, the area of the Suaeda marsh in human management areas and natural conditions will be3.26times and5.46times compared with scenario I. If removal the scenario of the Spartina mash, the area of the Suaeda marsh are7.24times and20.65times compared with scenario I in human management areas and natural conditions, respectively.更多还原