【作者】 李良涛；

【导师】 宇振荣；

【作者基本信息】 中国农业大学 ， 生态学， 2014， 博士

【摘要】 庭园和农田边界分别作为农业/农村生态系统中乡村聚落景观和农业生产景观中生物多样性保护的热点区,对这些生境植被的合理管护能有效提高地方生物多样性、生态系统服务功能,增进人类福祉。通过田野调查、访谈,分析并探讨农业景观中农田边界、田间岛屿等非农生境和农村居民庭园植被物种多样性、功能组成及其立地、景观和管理等关键因素影响,并提出了生态植被营建途径及管护建议,为加强农村/农业生态系统管护、优化生态服务提供参考。1、以华北平原集约化农业景观为研究区,对14个典型生境的植物组成进行分层随机调查,结果显示：(1)在样地水平上,河边坑塘植被带物种丰富度最高,其次为沟渠,果园、棉田、粮田物种丰富度较低,与菜地等4个农业生产生境草本乡土物种比例均≤70%,其中菜地仅30.5%。(2)在生境水平上,田间坟地(9-800m2)拥有最高的物种丰富度(102种)、草本物种(85种)和乡土特有种(6种),林地中有最多的木本物种(13种),其他生境均有特有种。(3)在所有生境中,共计调查到植物161种,分属41科,119属,其中野生种117种,草本131种,草本中58%为乡土物种,74%为杂草,22.1%为恶性杂草；其中野生濒危保护植物3种、外来入侵种(潜在)15种。对于不同尺度,样地尺度比生境尺度上拥有较低的乡土植物和较高比例的杂草；对于不同生境,物种盖度聚类和排序分析显示,田间坟地植被与荒草地不同,具有明显近自然特征,粮棉果蔬农业生产生境与廊道及田间岛屿等半自然生境间差异较大。(4)景观尺度研究表明,从单作农业景观、复合种植农业景观至非农生境镶嵌异质景观等梯度不同的五个景观类型区之间植物丰富度和组成具有显著差异；当农业生产生境饱和后,外来物种和恶性杂草增长趋缓,而随着非农生境多样化,多年生杂草、木本等逐渐增加,尤其是乡土种递增最为明显。2、系统调查了506条典型农业景观中河渠、生产路、田间路、田间草埂、树埂等5类线性非农生境的植物群落组成。研究显示：共计调查到植物物种123种,农田边界植被科优势现象比较突出,单种属较多。农田边界自生植被型以灌草丛为主,乔木层多为人工栽植,且较单一集中,群落物种组成多以野生杂草种为主。尽管很多物种为广布种,不同类型农田边界在植被物种丰富度、物种功能组成、植被分层及盖度上差异较大。(3)样地尺度α多样性河渠最高,田间草埂最低,β多样性和γ多样性田间路最高,生产路最低。(4)不同农田边界类型之间在群落组成上也有明显差异,其中生产路和田间草埂相似度最大,生产路和田间树埂相似度最低。3、对农田边界植被组成及其立地和景观尺度等一系列解释变量的研究显示：(1)农田边界植物多样性及其组成受到立地和景观不同尺度众多因素的显著影响；(2)除邻近土地利用、植被结构、样地环境影响因素外,景观特征和管理因素也不容忽视；(3)不同的农田边界功能群物种组成不同,河道沟渠和田间树埂具有较高的自然种、木本种及多年生草本多样性,而农业耐受种较低,含有乔灌层的农田边界拥有较少的农业耐受种和一年生草本,较宽的田间路两种功能组植物丰富度都较高,而较窄的生产路和农田草埂农业耐受种和一二年草本较多。(4)用PCCA方法分析5个不同变量组的影响,结果显示,除受土地利用因素、植被结构和立地环境等立地因素主要影响外,农田边界植物多样性及组成受到的景观因素的影响比管理因素大。4、以集约化农业景观为研究区,基于乔、灌、草组合模式划分的5种植被结构类型,对“田间岛屿”的植物多样性及其历史和面积因素影响进行了研究。结果显示：(1)在10个1×1km2景观栅格中,共计调查到494个田间岛屿,平均田间岛屿斑块面积为68.3m2,总面积3.38ha,占研究区总面积的0.36%,丰富度却占到景观总物种的64%,是维系该地区农业景观生物多样性重要的热点区域和物种库：(2)田间岛屿植被乡土物种占主导地位,每种植被结构类型均有特有种,调查中记录了3种重点保护濒危植物；(3)不同类型植被结构α、β、γ多样性差异明显,其中物种丰富度从植被结构类型1至5,依次降低,类型4、5有较高的beta多样性和较低的γ多样性；(4)线性回归分析显示,岛屿历史和面积两个因素对样地物种丰富度影响明显,对乔木层和灌木层盖度影响较大,而对草本层无明显影响,研究强调了“田间岛屿”这一非农生境对于维系和恢复集约化农业景观生物多样性的重要性。(5)对田间岛屿生境地被自生植被组成的立地和景观影响因素研究结果显示,在记录的28个环境变量中,有15个显著影响变量参与模型建立,PCCA结果表明植被群落结构因子和景观因子的影响都大于立地环境的作用。5、对北京沿城市化梯度农村居民庭园植被研究显示,沿该梯度庭园的物种多样性和丰度格局反映了不同地区居民的需求层次差异。从北京近郊到远郊村庄,庭园生物多样性和丰富度的变化呈现从观赏(文化服务)到食用(供给功能)的变化。Rarefaction曲线显示观赏物种是导致地区间庭园p多样性差异的重要因素：排序也显示了物种组成沿城市梯度的显著差异,近郊和偏远农村差异最大,而郊区则与两者均有较高相似度。此外,物种与庭园面积仅在偏远农村呈正相关。较高的食用植物盖度和物种密度,显示了农村庭园种植者比城市的有较高的食用功用需求。6、对北京农村居民庭园植被垂直配置和水平微环境组成进行定量化研究,分析了庭园面积、不同村庄及城乡梯度对庭园结构的影响。在105户居民庭园中,43.8%具有庭园外园地,农户宅基地面积较小的户倾向于种植外园地,庭园(内外园地)面积平均为189.2ma(37-746m)9；所有庭园中共调查到17种土地利用类型,单个庭园土地利用多样性3-14个不等平均8.9个；不同城乡区位会影响庭园水平土地利用多样性,近郊庭园较少,而且也会影响其面积差异,较大户级总面积支持了较多水平土地利用多样性,有的类型分布比较固定；庭园各植被层盖度和>3m的乔木棵数均随着庭园面积增加而递增,城乡区位也会影响居民庭园不同植被层盖度及所占总面积比例,<3m各植被层面积,偏远乡村较大,>3m植被冠层盖度在中间梯度的远郊庭园面积最大。7、应用多元回归方法定量化研究预测了庭园内外部环境、庭园管理及农户社会经济特征对庭园植物组成、植物多样性及庭园功能类型的影响,结果显示：(1)庭园面积和家庭收入、庭园结构和庭园管理投入时间等因素是影响的重要因素；同时,农业占家庭经济的比重增加,户主年龄较年长、受教育水平较低、家中老人较多,食用物种丰富度也较高；户主年龄较年长、受教育水平较低、家中老人较多,乡土物种较多；庭园历史较长、庭园管理者有男性参与提高了庭园中遮阴、木材等其他功用植物丰富度；户主为女性促进了庭园总物种丰富度、观赏和草本种丰富度的增加。(2)花园型分布与户主为女性显著相关；距离城郊越远、户主从事农业为主,农业收入农业比重大,庭园结构复杂、庭园管理投入时间多、庭园种植历史较长、主要管理者有男性参与,对菜园分布有积极影响；林园型分布主要与庭院面积和家庭收入有关,庭院面积较大、家庭收入较高,林园分布的比例就较高；距离城市较近、户主从事非农职业,庭园结构简单、较少对庭园进行管理促使过度铺装庭园的分布概率增加；庭园结构复杂、管理时间较多、庭园历史较长、管理者有男性参与对多功能复合型的分布有积极影响。8对农业景观不同类型生境进行良好的管护需要在适当的地方采用合理的措施,实施正确的方法进行生态植被营建和管护,参考国内相关研究和实践案例,对农业景观中非农生境和庭园生态植被提出管护建议和建设注意事项,使人们对农村多种生境更加熟悉,并应用于设计和实践。更多还原

【Abstract】 Agricultural landscape and rural settlements depend on hotspots of biodiversity and ecosystem services that are found in field margins and homegardens.Indeed, these habitats can be managed and designed to host local biodiversity, with neutral positive effects on ecosystem services and human well-being. Based on fieldwork, interviews, variation in species richness, function composition and habitat specificity among field margins, infield islets and homegarden was explored, in combination with variable, including habitat variability, landscape variable, management and households’contextual attributes.In order to strengthen agricultural and rural eco-environments management, maximize ecosystem service, then we briefly outline the principal approaches and measures for vegetation construction and management within agricultural-rural ecosystem.1In this study the plant species diversity, composition and abundance is examined at the scale of site, habitat and landscape for14farm habitats in the North China Plain by stratified random sampling. The result show that (1) at the site level, river vegetation had the highest richness of overall, followed by ditch. Orchards, cereals, vegetable, orchard had few species and native richness (<70%), especially for vegetable field only35.5%.(2) At the site level, Sacred graves as infield islet (9-800m2) had the highest species richness in total (102), herbs (85) and spontaneous native unique species in one specific habitat(6). Woodlots had the most woody species (56). Other habitats had spontaneous unique native species also.(3) For all habitats,161species belonging to119genera and41families were recorded.The spontaneous species recorded represented81species, whereas herbaceous represented131,58%of which were native,74%were weeds, and22.1%were noxious weeds. In this survey,3wild endangered species and15potentially invasive species were recorded. Cotton fields, field road (≥4m) and river (pond) have the most frequent of it. Site scale has lower percent of native species and higher proportion of weeds than habitat scales. The cluster and DCA analysis for herbs showed that sacred grave with nearly natural features and grass were different, as were cotton fields, cereals, vegetable, orchard and noncrop habitats.(4) Landscape level study revealed that species composition and diversity varied significantly among the five landscape scenarios, which from a monotonous agricultural landscape to a diverse mosaic of crop and non-crop composite landscape. When all agricultural production habitats appeared, growth of alien species richness and noxious weeds became to flatten, and perennial weeds, woody, particularly native species became to increased steeply for joining non-agricultural habitats.2The vegetation compositions of506individual field margins were inventoried systematically. The results of this study are as follows:(1) a total of123species were recorded in this investigation. The dominant families are prominent. The single species genera are rich.(2) The spontaneous vegetation is dominant by grasses, herbs and shrub. Planting trees are dominant by single-species. Despite having many widespread species, vegetation structure, species richness, function composition and species cover differed between the five field margin types.(3) The rivers&ditches have the highest a diversity and the herbaceous field ridges have the lowest one. The roads (width>4m) have the highest (β-diversity and y-diversity. And the field roads (width<4m) have the lowest ones.(4) Community similarity analysis shows that the field roads (width<4m) and the herbaceous field ridges have highest inter-community similarity, while field roads (width<4m) and the wooded field ridges have lowest one.3The result of field margins showed that (1) the plant species richness and composition of field margins were affected significantly by most explanatory variables including local and the landscape factor. Multivariate analysis showed that (2) besides adjacent land use, structure, abiotic environmental conditions, the role of other factors such as landscape factors and management could not be neglected.(3) The type of field margin had a special impact on vegetation compositions—riverside&ditches and woody field ridge enhanced the richness of nature-value species, woody species, perennial herbaceous and suppressed agrotolerant species, annual&biennial species; boundaries with tree or shrub layer hosted only a few agrotolerant, annual and biennial species. Wider road verges increased the species richness of both plant functional groups while narrow road verges hosted more agrotolerant, annual and biennial species.(4) PCCA methods revealed the ranking of the5subsets. Besides a predominant impact of land use, structure, abiotic environmental, it was found that landscape context have a relatively larger influence on the distribution of plant species in field margins than management.4This study analyzed the distribution of plant species richness and the impact of site history and islet area on plant diversity in small remnant habitats midfield islets in101×1km2agricultural landscapes grid cells in Quzhou, Hebei Province. Five habitats included different types of vegetation: deciduous tree, shrub, mixed herbaceous and herbaceous layers. The result showed that (1) there were494midfield islets in the study agriculture landscape and103species were found on200investigated midfield islets. The mean area of the islets was68.3m2. Although they occupied a relatively small area (3.38ha) the majority (64%) of total species in the study agriculture landscape were found in the islets.(2) From the characteristics of the plant species traits, the native was the leading one. Different types of vegetation have habitat specialists which species restricted to only one habitat. In this survey,3wild endangered species were recorded.(3) Alpha diversity was highest in type1and β diversity was highest in type4and type5habitats. γ diversity was highest in type2and lowest in type5.(4) Line regression shows that diversity is positively related to area and age of midfield indicating a strong history and area effect. Area and history are also shown to have an independent effect on tree and shrub covers. Our study suggests that maintaining non-crop habitats such as midfield islets is critical for sustaining and restoring biodiversity in intensive agricultural landscape.(5)Then the hypothesis that plant diversity in infield islets can be influenced by a range of variables including local and the landscape factor was tested. The explained variation in species richness was significant for15of the28variables. For the fraction of variable explained, STR and LDS are all more than SIT.5The gardens were distributed across three urbanized regions (suburban, peri-urban, and exurban). We found that species biodiversity and abundance shift according to a hierarchy of need. Garden biodiversity and abundance shifted from ornamentals (cultural ES) to edibles (provisioning ES) with distance from Beijing. Rarefaction curves indicate ornamental species drive (3diversity. Ordination also showed a shift in species composition across a distance gradient from the city; Suburban and exurban gardens were the most different, while peri-urban gardens were similar to both. High edible cover and high species density indicates that demand for edibles in exurban may be higher than suburban.6We present the detailed analysis of variation in vegetation structure and land-use of home gardens, in relation to garden area and urbanization, across five villages in Beijing. In all105homegardens studied,43.8%with smaller parcel area have a border garden. The average garden size including inside and outside for gardening is189.2m2(37-746m2).The land-use richness per garden ranged from3to14across all five villages with an overall mean of8.9.Garden size and urbanization gradient had significant influence on garden configuration. Larger size supported more land-use richness. The land-use richness of suburban was significantly lower. Some features have a relatively fixed distribution. The area of various vegetation layers and the number of trees>3m in height in gardens increased significantly with garden area. The effect of urbanization gradient on vertical vegetation area and the proportional contribution of garden were also significant. For the cover of vegetation layers of<3m, the exurban is the largest. But for>3m, peri-urban is the largest. 7This research investigated whether garden biophysical variables, management and households’ characteristics have a quantitative and predictable relationship with homegardens vegetation composition, plant diversity and garden type using multiple regression methods. The result reveal that:(1) garden size and family income, land-use within gardens and vertical vegetation structure significantly affected plant diversity of homegarden. At same time, with the increase of agricultural in the proportion of family income, the age of household head, old persons, and the decrease of education years of household head, the edible species richness is higher. The age of household head and education were associated with the native plant richness. Garden history and gender of garden labor were associated with shelter and wood species. And female household head promoted overall species richness, ornamental and herbaceous species richness.(2) Female household head had the strongest relationship with ornamental garden.Garden size and family income were strong predictors for the woody garden. Distance to urban, major source of family income, garden structure, and labor invested for homegardening have a predictable relationship with paving over garden. Distance to urban, major source of family income, agricultural in the proportion of family income, garden structure, labor invested, garden history and gender of labor for gardening were strong predictors of edible homegarden. Garden structure, labor invested, garden history, and gender of labor for gardening was strong predictors of multifunction homegarden.8Good vegetation stewardship on the agricultural landscape is all about choosing the right measures, putting them in the right habitats, and managing them in the right way. We briefly outline the principal approaches and measures for vegetation construction and management within agricultural-rural ecosystem based on our research and previous experience. These suggestions is intended to help anyone undertaking subdivision in the rural environment achieve good practice and design.更多还原