【作者】 杨予海；

【导师】 刘荣堂；

【作者基本信息】 甘肃农业大学 ， 草业科学， 2007， 博士

【摘要】 通过对青海省三角城种羊场草地农业生态系统中的气候、土壤、植被、植被光合效率、能流以及生态经济效益等方面进行了全面调查,结合TM影像资料,利用遥感及地理信息系统软件,对青海省三角城种羊场草地类型及其景观动态变化进行了研究,涉及遥感影像分析、地理数据库建模、草地类型波谱分析、草地资源调查等,进行了系统分析研究,取得了以下主要研究成果:青海省三角城种羊场草地类型主要包括高寒草原、高寒草甸、高寒荒漠、沼泽化草甸等,其中高寒草原和高寒草甸所占的面积最大,在畜牧业经济和生态保护中也最为重要。植被分布和土壤吸附水状况,代表不同草地型的优势种和亚优势种植物随着海拔高度上升,依次为海韭菜、鹅绒萎陵菜→芨芨草、羊茅→冷地早熟禾、紫花针茅→矮嵩草、小嵩草,表明该羊场草地型垂直分布特征明显;海拔越高,植物高度和地上生物量越低,说明高海拔的环境条件对草地生产能力的强硬制约性;不同海拔高度,不同土壤深度土壤吸附水达到等吸附量点的时间不同,海拔越高,达到等吸附量点的时间越早,证明同一地区的吸附水含量由于海拔差异而相差很大。在中度到重度退化进程中,草地植物群落随着退化程度的加重,其物种多样性呈减小的趋势,而建群种的重要值呈增大趋势。针茅型中度与重度退化样地植物群落物种多样性指数(a多样性指数,下同)分别为2.50、2.26,建群种的重要值分别为19.09、23.8。嵩草型中度与重度退化样地植物群落物种多样性指数分别为2.61、2.42,建群种的重要值分别为13.31、22.81。针茅型与嵩草型不同退化样地地上植物类群及总植物量季节动态均呈现明显的“单峰”曲线,其峰值同时出现在8月下旬,在植物生长季,主要植物类群地上植物量季节动态具有明显的差异。两草地型地上植物量季节积累动态曲线可用y=ea+bx(生长曲线)很好拟合。随着草地退化程度加重,针茅型样地禾草类在草群中所占比例下降(以植物干重为测度),杂草类在草群中所占的比例上升。群落干物质积累速率下降,最终导致草地净初级生产力的降低。嵩草型中度退化样地的净初级生产力高于重度退化样地,但由于在牧草返青期放牧使中度退化样地禾草类在草群中所占的比例小于重度退化样地。针茅型中度与重度退化样地地上植物量净生产量分别为104.99g/m2.a、48.38g/m2.a,下降达53.92%;嵩草型中度与重度退化样地地上净生产量分别为144.82g/m2.a、87.96g/m2.a,下降达39.26%。针茅型与嵩草型样地地下80%以上的植物量分布在0～20cm的土层中,用y=ax+b能够很好描述两草地型在生长季各土层中的平均植物量及其所占比例。在地下40cm深的土层中,针茅型中度与重度退化样地地下植物量的净生产量及周转值分别为849.6g/m2.a、32.45%、1346.05g/m2.a、51.93%;嵩草型中度与重度退化样地地下植物量的净生产量及周转值分别为787.13g/m2.a、26.08%、1190.6g/m2.a、51.33%。地下与地上植物量的比值随着草地退化程度加重而增大。针茅型中度与重度退化样地地下与地上植物量的比值分别为16.86、51.51;嵩草型中度与重度退化样地地下与地上植物量的比值分别为24.61、27.22。针茅型与嵩草型不同退化样地热值测定表明,群落地上部分热值大于地下部分,地上部分热值含量以牧草生长旺盛期最高,返青期最低,枯黄期次之。并且其热值含量均高于世界陆生植物的平均热值,地上、地下、全群落能量积累与植物量增长呈显著的线性正相关。草地净能空间分配率随着草地退化程度加重而减小。针茅型中度与重度退化样地分别为0.1667、0.0439;嵩草型中度与重度退化样地分别为0.2257、0.1049。随着草地退化程度的加重,植物群落地上部分对光能的转化效率减弱。两草地型对光能的转化效率:全群落,针茅型中度与重度退化样地为0.208%、0.317%,嵩草型中度与重度退化样地为0.216%、0.259%;地上部分,针茅型中度与重度退化样地分别为0.03%、0.013%;嵩草型中度与重度退化样地分别为0.04%、0.024%,地下部分,针茅型中度与重度退化样地分别为0.178%、0.303%,嵩草型中度与重度退化样地分别为0.177%、0.234%。在分析青海省三角城种羊场生态系统的自然、社会经济概况的基础上,运用能值理论与方法,通过对青海省三角城种羊场2002-2006年的净能值产出率(NEYR)、能值投入率(EIR)、环境负载率(ELR)和可持续发展能值指数(ESI)等指标的计算及趋势分析得出:青海省三角城种羊场生态系统投入的能值总量为5.00×1029sej,其中可更新环境资源、不可更新环境资源、不可更新工业辅助能及可更新有机能分别占系统输入能值总量的7.39%、7.68%、35.15%和49.77%。环境资源总投入占总能值投入的15.08%。在辅助能投入中,可更新的有机能占辅助能投入总量的58.61%。在环境资源总投入中,不可更新环境资源投入占其中的50.96%。青海省三角城种羊场能值投入率(EIR)值由2002年的0.0489缓慢增长至2006年的0.734,总体在0.05～0.07间浮动。青海省三角城种羊场能值产出率(EYR)由2002年的57.4上升至2005年的83.7,总体上处于波动上升态势,表明青海省三角城种羊场经济活动的能源利用效率在逐年提高,竞争力明显增强。青海省三角城种羊场能值投入率(EIR)值由2002年的0.0489缓慢增长至2006年的0.734,总体在0.05～0.07间浮动。这表明青海省三角城种羊场经济对外开放和利用外界各类“资源”程度较低。青海省三角城种羊场环境负载率(ELR)由2002年的31.3增至2006年的71.4,总体呈波动上升趋势,表明青海省三角城种羊场经济发展对其环境系统的压力逐步增加。青海省三角城种羊场能值—货币比(EDR)呈逐年上升趋势,由2002年的1.50×1014sej/$上升至2006年的7.98×1014sej/$,这主要是青海省三角城种羊场生产总值在2002—2006年间的增长缓慢,其经济系统的开发程度降低造成的。2002-2006年青海省三角城种羊场ESI值由58.46降至21.8,呈波动下降趋势,同期青海省三角城种羊场ELR由31.3增至71.4呈快速上升趋势。由ESI=NEYR/ELR等式可知,ESI的下降是由于2002-2006年青海省三角城种羊场环境负载率(ELR)的增长速率快于同期净能值产出率(NEYR)的增长速率。不同地被在TM影像上各波段的反射特征。通过对石山冰川、盐碱地、水库、裸地、高山草原、高寒草甸、河流、人工草地、湖泊和沼泽地10种不同地被类型及山地草原、山地草甸和高山草甸三种植被类型的波谱曲线调查研究结果发现:在TM1、TM2和TM3可见光波段,湖泊和沼泽地的反射率远远高于其它地被;而三种草地植被在可见光区的反射率差异不大,到红外光区,它们的反差才急剧加大。在TM4近红外波段,绿色植被反射率差异较大,为区分植被的理想波段。在TM5波段,高寒草甸、裸地和高寒草原的反射率远远高于其它地被。在TM6波段,各地被的反射率差异不大。不同土壤类型的土壤线。根据土壤在红外波段和近红外波段反射率之间的线性关系,将它们的相关性线性回归得到四种土壤类型的土壤线依次为:沼泽土:NIR=0.46R +13.91,相关性0.83;暗栗钙土:NIR = 0.44 R+22.66,相关性0.78;高山草原土:NIR = 0.52R +12.62,相关性0.80;高山草甸土:NIR =0.57 R + 6.91,相关性0.90。土壤线的斜率大小排列顺序为:高山草甸土>高山草原土>沼泽土>暗栗钙土,而截距为高山草甸土<高山草原土<沼泽土<暗栗钙土。对ETM遥感影像各个波段分析结果表明1、2、3波段间彼此高度相关, 4、5、7波段间有一定相关性;在波段7、5、4、3处,草地类型波谱曲线变化幅度大,宜用作草地分类;利用数字高程模型,辅助草地生态序列分布规律,结合遥感光谱特征进行了草地类型,对解决草地类型间因光谱重叠而难以区分的难题找到了可行方法。依据RVI和NDVI两个植被指数对植被生长状况与草地产草量有不同的敏感度,采用微积分方法,化整为零,按盖度区间,计算植被指数同产草量的线性方程,使遥感预测产量与草地实际产量的相关性更好。对青海省三角城种羊场景观生态系统的系统耦合结果发现,系统耦合可根据干扰性质分为自然耦合和人为耦合。自然耦合是人为耦合的基础,系统生态生产力的提高取决于人为耦合的优化过程。更多还原

【Abstract】 The grassland types and their landscape dynamic of pastoral-agriculture ecological system in Sanjiaochen Sheep Breeding Farm Qinghai Province were studied through the thorough investigation. The data about climate, soil, vegetation, conversion efficiency of solar radiation, energy storage as well as the ecological benefit were collected. The technologies applied in this research were TM image, remote sensing and geographic information system(GIS). The images taken by the remote sensor were analyzed, the modeling of the geographic database was built, the characters of the spectrum of the grasslands were analyzed and the investigations of the grassland resources were conducted. The data systematic analysis presented the results as following:Grassland types in Sanjiaocheng Sheep Breeding Farm mainly are Alpine steppe、Alpine meadow、desert steppe etc. Among them Alpine steppe and Alpine meadow not only occupy the largest area compare to other types but also play an important role in livestock production as well as ecological protection. In terms of vegetation distribution characteristics and soil hygroscopic water conditions, dominant species and sub-dominant species that represent different grassland types in different altitudes ( from 3100m to 3700m ) are in order as follows:Triglochin maritimum , Potentilla anserine→Achnatherum splendens,Festuca ovina→Poa crvmophila,Stipa purpurea→Kobresia humilis,Kobresia humilis. Vertical distribution zonation of grassland type was clearly showed here; higher the elevation is, lower plant height and aboveground biomass are, it is certified that fierce environment condition in high altitude puts special stress on grassland productivity. The equilibrium of hygroscopic water in different soil level is not reached simultaneously at different altitude. Higher the elevation is,earlier this equilibrium is reached. It is proved that there are remarkable differences in hygroscopic water content within a small area due to elevation events.The species diversity(a diversity index)of plant community decreased and important value increased with the aggravation of degraded degree. The diversity indexes of Stipa purpurea community with moderate and heavy degradation were 2.50、2.26 respectively,the important value of dominant species were 19.09、23.8 respectively. The diversity indexes of Kobresia capillifolia community with moderate and heavy degradation were 2.6069,2.4167 respectively,the important value of dominant species were 13.31、22.81 respectively. Season dynamic of abovegroung population phytomass of Stipa purpurea and Kobresia capillifolia grassland types take on single apex curve,the value of apex appeared in the last ten day of August,the phytomass of aboveground population of two grassland types have clear difference. The equation y=ea+bx simulate season dynamic curve of aboveground phytomass of Stipa purpurea and Kobresia capillifolia grassland type well.The dry weight ratio of the grasses of Stipa purpurea type to community decreased but that of forbs increased with the aggravation of degradation. The rate of dry matter accumulation decreased,which finally lead to decreasing net primary production. The net production of above-ground phytomass of moderate-degraded plot of Kobresia capillifolia type was more than heavy-degraded plot,because of grazing in germination,the grasses percentage of moderate degraded plot of Kobresia capillifolia type was less than heavy-degraed plot. The net production of above-ground phytomass of Stipa purpurea community with moderate and heavy degradation were 104.99g/m2.a、48.38 g/m2.a respectively. The net producion of above-ground phytomass of Kobresia capillifolia community with moderate and heavy degradation were 144.82g/m2.a、87.96g/m2.a respectively. The more than 80% belowground phytomass distributed at 0～20cm soil layer,the equation y=ax+b simulate mean phytomass and its percentage of individually soil layer at growing period. Under the 40 centimeter deep soil layer,the net production and turnover value of belowground-ground phytomass of Stipa purpurea community with moderate and heavy degeneration were 849.6g/m2.a、32.45%、1346.05g/m2.a、51.93% respectively. Those of below-ground phytomass of Kobresia capillifolia with moderate and heavy degradation were 787.13 g/m2.a、26.08%、1190.6g/m2.a、51.33% respectively.The ratio of below-ground phytomass to above-ground phytomass increased with the aggravation of degradation. The ratios of below-ground phytomass to above-ground phytomass of Stipa purpurea community with moderate and heavy degradation were 16.86、51.51 respectively. The ratios of below-ground phytomass to above-ground phytomass of Kobresia capillifolia community with moderate and heavy degradation were 24.61、27.22 respectively.The measures of caloric value of every plot showed that:The caloric value of above-ground more than that of below-ground,the caloric value of above-ground was the highest on the forage flourishing stage,the lowest on green up period,on the withering stage in the second level. the caloric value of above-ground more than that of the terraneous. The energy storage of above-ground,under-ground and total community were linear positive significance to the biomass. The net production of spatial distribution decreased with the aggravation of degraded degree. The net production of spatial distribution of Stipa purpurea community with moderate and heavy degradation were 0.1667、0.0439 respectively,that of Kobresia capillifolia community with moderate and heavy degradation were 0.2257、0.1049 respectively.The conversion efficiency of solar radiation of above-ground decreased with the aggravation of degradation.The conversion efficiency of solar radiation of above-ground of Stipa purpurea community with moderate and heavy degradation were 0.03%、0.013% respectively , that of below-ground were 0.178%、0.303% respectively,that of total community were 0.208%、0.317% respectively. The conversion efficiency of solar radiation of above-ground of Kobresia setchwaneusis community with moderate and heavy degeneration were 0.04%、0.024%,that of below-ground were 0.177%、0.234% respectively,that of total community were 0.216%、0.259%.Based on the analysis of nature and social economy status of SanJiaocheng sheep breeding farm eco-system in Qinghai Province,the emergy theory and method were applied to calculate and analyze the NEYR,RIR,ELR and ESI of the SanJiaocheng sheep breeding farm during the year 2002-2006: The total emergy quantity input on the San Jiaocheng sheep breeding farm was 5.00×1029sej, including renewable environment resources、non-renewable environment resources、non-renewable industry attached energy and renewable organic energy which occupied the total input emergy of 7.39%、7.68%、35.15% and 49.77% respectively. The input environment resources occupied 15.08% of the total emergy. In the input attached energy,the renewable organic energy occupied 58.61% of the total input attached energy. In the input environment resources,the non-renewable environment resources occupied 50.96% of the total input environment resources. The EIR of San Jiaocheng sheep breeding farm in Qinghai Province increased from 0.0489 in 2002 to 0.734 in 2006 gradually,and the overall EIR fluctuates between 0.05 and 0.07. The EYR of SanJiaocheng sheep breeding farm in Qinghai Province increased from 57.4 in 2002 to 83.7 in 2005,and the overall tendency of EYR was on the rise,which indicated that the energy efficiency of economic activity on Sanjiaocheng sheep breeding farm in Qinghai Province increased year by year,and the competitive ability obviously enhanced. The EIR of SanJiaocheng sheep breeding farm in Qinghai Province increased from 0.0489 in 2002 to 0.734 in 2006 gradually,and the overall EIR fluctuates between 0.05 and 0.07,which indicated that the degree of economy opening to the outside world and the utilization of various outside“resources”was low in that area. The ELR of SanJiaocheng sheep breeding farm in Qinghai Province increased from 31.3 in 2002 to 71.4 in 2006, and the general tendency was on the rise,which indicated that the pressure of economic development on the environmental system increased gradually in that area. The EDR of SanJiaocheng sheep breeding farm in Qinghai Province showed a tendency of rising year by year,it increased from 1.50×1014sej/$ in 2002 to 7.98×1014sej/$ in 2006,this mainly because that the gross agricultural production of that area increased slowly during the year 2002-2006 and the decreased exploiting degree of economic system. During the year 2002-2006 the ESI of SanJiaocheng sheep breeding farm in Qinghai Province decreased from 58.46 to 21.8,the general tendency was on the decrease,and the ELR increased from 31.3 to 71.4 at the same period,which showed a rapid rising tendency. From the equation ESI=NEYR/ELR thus it can be seen that the decrease of ESI during the year 2002-2006 was due to the growth rate of ELR was higher than that of NEYR at the same period on the SanJiaocheng sheep breeding farm in Qinghai Province. Reflection characteristics of TM image bands for different objects. Ten objects were chosen for their band comparison,including rocky mountain glacier,saline and alkaline area,reservoir,bare land, alpine steppe,alpine,meadow,river,artificial grassland,lake and swamp. Three vegetation types were also selected for plant band comparison,such as alpine steppe,alpine meadow and mountain meadow. The results showed: among TM1,TM2,and TM3 bands,spectral reflectance of bog and lake is much higher than that of other objects. There is no significant difference among these three types of vegetation. While within band 4,spectral reflectance of green plants is different from each other,therefore,band 4 is good at distinguishing different vegetation. As for band 5,spectral reflectance of alpine meadow,bare land and alpine grassland is much higher than other objects. As for band 6,there is no distinctive variation for all objects.Soil lines of different soil types. According to linear relationship between spectral reflectance of infrared band and near-infrared band,their correlative relationships were linear regressed to get soil lines of 4 soil types: Swamp: NIR=0.46R +13.91,relevance 0.83; Dark chestnut soil: NIR = 0.44 R+22.66, relevance 0.78; Mountain steppe soil: NIR = 0.52R +12.62, relevance 0.80; Mountain meadow soil: NIR =0.57R + 6.91,relevance 0.90. As we can see from the comparison of different soil lines,slope: alpine meadow soil>alpine steppe soil>swamp soil>dark chestnut soil; intercept: alpine meadow soil<alpine steppe soil<swamp soil<dark chestnut soil.An analysis on each band of the ETM image was made,result showed that band 1,2 and 3 were related to each other greatly,band 4,5 and 7 were only related in some extent. The spectral curves of different grasslands changed greatly near band 7,5,4 and 3,which could be infered that these bands could be used in the classification of the grasslands. The type of grassland changed with the changing of the topographic factor in a regular order,based on the theory,this experiment used the DEM assistanted with the distribution of the ecological sequence of the grassland,combined with the characters of the spectrum of the RS to plot out the grasslands,which provided a possible way to solve the problem that it was difficult to plot out different grasslands because of the overlapping of the spectrum.With the changing of the coverage of the vegetation,the sensitivity of RVI and NDVI to the productivity of the grassland and the growing status of the vegetation was different. The calculous mathematics was adopted to break up the whole into parts to compute the linear equation between the vegetation index and the productivity according to different ranges of coverage,results showed that the relativity between the mathematical model and the real productivity of the grassland had become better than anyother methods ever used.A study on the coupling action was made about the landscape ecology of the San Jiaocheng sheep breeding farm,the result showed that the coupling action between the alp meadow and the mountains meadow was direct,it was same to the mountains meadow and the mountains campo,while it was indirect between the alp meadow and the mountains campo and the coupling action of them needed the medium of the roads or water system,or the indirect action of the herding behavior of the livestocks. The number of the grassland types in the mountains campo was more than that in any others’,and it was same to the naked spot. The coverage and the productivity of the grassland were lower,which means that the grassland in the mountains campo was in the stage of degeneration. The coupling of the system could be divied into natural coupling and artificial coupling,at the same time,the natural coupling was the base of the artificial coupling,and the improvement of the ecological productivity of the system depended on the optimized process of the aritificial coupling.更多还原