共查询到20条相似文献，搜索用时 31 毫秒
1.
Jennifer L. Hopwood 《Biological conservation》2008,141(10):2632-2640
Marginal habitats such as hedgerows or roadsides become especially important for the conservation of biodiversity in highly modified landscapes. With concerns of a global pollination crisis, there is a need for improving pollinator habitat. Roadsides restored to native prairie vegetation may provide valuable habitat to bees, the most important group of pollinators. Such roadsides support a variety of pollen and nectar sources and unlike agricultural fields, are unplowed, and therefore can provide potential nesting sites for ground-nesting bees. To examine potential effects of roadside restoration, bee communities were sampled via aerial netting and pan trapping along roadside prairie restorations as well as roadsides dominated by non-native plants. Management of roadside vegetation via the planting of native species profoundly affected bee communities. Restored roadsides supported significantly greater bee abundances as well as higher species richness compared to weedy roadsides. Floral species richness, floral abundance, and percentage of bare ground were the factors that led to greater bee abundance and bee species richness along restored roadsides. Traffic and width of roadside did not significantly influence bees, suggesting that even relatively narrow verges near heavy traffic could provide valuable habitat to bees. Restored and weedy roadside bee communities were similar to the prairie remnant, but the prairie remnant was more similar in bee richness and abundance to restored roadsides. Restoring additional roadsides to native vegetation could benefit pollinator conservation efforts by improving habitat on the millions of acres of land devoted to roadsides worldwide, land that is already set aside from further development. 相似文献
2.
Recent declines in North American honeybee populations have highlighted the importance of native bee conservation, and the need for research on the ecological requirements of native bees in farmland. In this study, we investigated the value of hedgerows as foraging habitat for native bees in mosaics of small-scale agriculture and natural vegetation in two riparian landscapes in southeast Arizona, USA. In the summers of 2002 and 2003, we surveyed bees and flowers in four habitats: hedgerows, agricultural fields, woodlots, and native woodland. We asked: (1) How do hedgerows compare to other available habitats in bee abundance and species richness? (2) How does bee species composition in hedgerows compare to species composition in agricultural fields and woodland? (3) How do flower resources in hedgerows compare to those in fields and woodland?We found that hedgerows were attractive foraging habitat for native bees, especially in early summer, when hedgerows tended to have higher species richness than other agricultural or natural habitats. Cumulative species richness was highest in agricultural fields, although cumulative species richness did not significantly differ among fields, hedgerows, and woodland. While bee faunas overlapped among habitats, bee assemblages in hedgerows were more similar to those in woodland than to those in fields. The hedgerow herbaceous flora was roughly intermediate to that of fields and woodland; hedgerows also supported high densities of woodland-characteristic shrubs. These flowering shrubs were important in attracting bees that were otherwise uncommon in the landscape, including some species that are potentially valuable pollinators of agricultural crops. 相似文献
3.
Sarah S. Greenleaf 《Biological conservation》2006,133(1):81-87
Pollination provided by bees enhances the production of many crops. However, the contribution of wild bees remains unmeasured for many crops, and the effects of anthropogenic change on many bee species are unstudied. We experimentally investigated how pollination by wild bees affects tomato production in northern California. We found that wild bees substantially increase the production of field-grown tomato, a crop generally considered self-pollinating. Surveys of the bee community on 14 organic fields that varied in proximity to natural habitat showed that the primary bee visitors, Anthophora urbana Cresson and Bombus vosnesenskii Radoszkowski, were affected differently by land management practices. B. vosnesenskii was found primarily on farms proximate to natural habitats, but neither proximity to natural habitat nor tomato floral abundance, temperature, or year explained variation in the visitation rates of A. urbana. Natural habitat appears to increase B. vosnesenskii populations and should be preserved near farms. Additional research is needed to determine how to maintain A. urbana. Species-specific differences in dependency on natural habitats underscore the importance of considering the natural histories of individual bee species when projecting population trends of pollinators and designing management plans for pollination services. Thus, to maintain an entire bee community, multiple approaches, including maintaining natural habitat, should be implemented. 相似文献
4.
Though it is undoubted that tropical bees are influenced by habitat composition, few studies have investigated the relative importance of both local and landscape-level habitat parameters in supporting large and diverse bee communities. The conservation of native bee communities within agroforestry landscapes is particularly urgent given the importance of pollination services within these systems. In this study, we examined tropical bee communities within a largely deforested shade coffee-growing region in Chiapas, Mexico. We used regression tree modeling to examine the response of bee functional groups to local and landscape-level habitat management. Our models revealed that the most predictive factors for bee abundance and species richness were the number of tree species, the number of tree species in flower, and the canopy cover of the local agroforestry landscape. Solitary bees were most abundant in habitats with high canopy cover, while social bees were most abundant in habitats with greater tree species richness. Cavity-nesting and wood-nesting bee abundance was positively affected by the amount of canopy cover in the farm, while ground-nesting bees were most abundant in habitats with a large number of tree species in flower. Our results demonstrate that across bee sociality groups, nesting guilds, and tribes, the most critical factor impacting native bee communities was within-farm local vegetation management. These results reveal the important role that agroforestry managers can have on biodiversity conservation, and the potential contribution they can make by creating resource-rich agricultural matrices. Specifically, our findings highlight the importance of diverse overstory tree management in supporting native bee communities within tropical agroforestry systems. 相似文献
5.
Pia E. Lentini Tara G. Martin Philip Gibbons Joern Fischer Saul A. Cunningham 《Biological conservation》2012,149(1):84-92
Pollination has received attention recently due to reported sharp declines of Apis mellifera in several locations, and it has been proposed that diverse native bee communities may be key for continued pollination of economically important crops. However, there is some inconsistency in the literature as to how these communities should best be managed. To address this issue, we collected bees from an intensively managed agricultural region in eastern Australia using blue vane traps. Both linear remnants of vegetation, which form part of a larger corridor network, and adjacent fields of native and exotic pastures, wheat, canola, and lucerne were sampled. A total of 3249 individual bees, representing four families and 36 species were collected. Highly modified environments of nectar-bearing crop supported the most species-rich bee assemblages, and the highest abundance of individual bee species. Distance from the remnants did not limit the body size of species occupying fields (up to 400 m). However, richness of bee assemblages also responded positively to the presence of conservation land in nearby areas, or the number of remnant native trees surrounding traps. Linear remnants of native vegetation contributed to assemblage heterogeneity by adding unique species to the regional pool. Our findings indicate that agricultural industries that currently rely on pollination by A. mellifera should ensure that intensive land use is complemented by untilled areas in the form of conservation land, or farm dams and scattered trees in fields, to support wild pollinators that may act as insurance against further future losses of managed hives. 相似文献
6.
北京昌平区农业景观野生蜂多样性的时空动态分布 总被引：1，自引：0，他引：1
近年来,由于生境质量的下降和生境丧失,野生蜂多样性急剧下降,严重威胁传粉服务和农业的可持续发展。为揭示野生蜂在农业景观镶嵌体中的时空分布格局及不同生境对野生蜂保护的重要性,本研究于2016年4—9月采用网捕法调查了北京昌平农业景观4种主要生境类型(人工林、自然灌木林、荒草地和桃园)中的野生蜂,分析不同生境类型中野生蜂多样性及其时间动态变化特征。结果表明:野生蜂及其中的大体长蜂和独居蜂的物种数和个体数,以及中体长蜂的个体数均在自然灌木林中最高,人工林中小体长蜂个体数最多,野生蜂总个体数和总物种数均在桃园中最低。在不同月份,所有功能群野生蜂均在自然或半自然生境中最高,5月桃园中最低,原因可能与桃园中对地表杂草的集约化管理有关。多度最大的4个优势物种依次为黄胸彩带蜂、铜色隧蜂、黄芦蜂和隧淡脉隧蜂,其在生境间的时间动态特征不同。调查的4种生境均为野生蜂提供了可利用资源,但自然生境和半自然生境的作用更大;其中,自然灌木林维持了较多具有较高潜在传粉效率的中体长和大体长野生蜂,具有更高的保护价值。为了促进该地区农业景观中不同功能群野生蜂多样性及传粉服务,需提高生境类型的多样性,同时在野生蜂活动高峰季节采取低集约化的生境管理方式。 相似文献
7.
滇东南南瓜传粉昆虫密度对生境丧失的差异性响应 总被引：2，自引：1，他引：1
响应多样性理论认为不同传粉蜂对环境胁迫的响应尺度和过程在时空上存在差异,然而目前尚不清楚我国重要农林生态系统传粉蜂对环境胁迫的响应规律。试验以云南省文山州广南县南瓜农田生态系统为研究对象,根据农田景观中自然半自然生境数量和面积的变化梯度和离自然半自然生境的距离梯度选择南瓜样地,抽样调查南瓜样地内熊蜂密度和蜜蜂密度。解译卫星影像获取土地类型数据,分析熊蜂密度和蜜蜂密度同自然半自然生境面积百分率和耕地面积百分率的相关系数,在最大相关系数对应的空间尺度下探索蜜蜂密度和熊蜂密度对生境丧失的响应规律。研究发现,熊蜂密度随自然半自然生境面积百分率的增加而显著增加,但蜜蜂密度并不受自然半自然生境面积的影响。同时,熊蜂密度随耕地面积百分率的增加而显著下降,但蜜蜂密度不受耕地面积的影响。此外,熊蜂密度随离自然半自然生境的距离增加而显著下降,但蜜蜂密度不受距离作用的影响。由此可见,云南省文山州南瓜传粉熊蜂和蜜蜂多度对生境丧失表现出多样的响应规律。熊蜂和蜜蜂对生境丧失的差异性响应可能是保障该区域南瓜受粉的重要机制。 相似文献
8.
近年来全球范围内蜜蜂多样性的下降引起了各国科学家与政府对与农业可持续生产密切相关的传粉服务的关注。为了认识和了解区域传粉服务的供给和需求状况,为制定区域传粉管理策略奠定基础,本研究在2016年对北京昌平7种常见生境的野生蜂多样性、植被组成、土壤紧实度、作物种植面积等数据进行了调查,根据野生蜂的物种组成、飞行距离、筑巢偏好和活动季节,生境的植物资源数据和筑巢适宜性数据,结合In VEST模型对研究区传粉服务的供给进行了评估;根据研究区传粉依赖作物种植面积和各种作物对传粉服务的依赖程度,对研究区传粉服务的需求进行了评估;通过叠加传粉服务供给和需求分级图,评估了研究区的传粉服务供需匹配状况。结果显示,自然林是最适宜野生蜂的生境,其次是人工林、荒草地和边界生境;果园既是野生蜂的生境,也对传粉服务有较高的需求;大棚几乎不供给传粉服务,但对传粉服务有较高需求。对传粉供需等级匹配的结果显示,昌平区供给等级高于需求等级的区域占34.2%,大部分分布于山区,有较高的传粉服务供给和较低的传粉服务需求,具有较高保护价值;供给等级低于需求等级的区域占13.9%,多分布于靠近山区的平原地区,虽然具有较高的传粉服务供给,但也具有最高的需求等级;供给等级与需求等级持平的区域占51.9%,远离山区的平原地区由于具有较多半自然生境,传粉服务的供需基本是匹配的。最后,文章对研究区传粉服务提升提出了管理建议:对于高供给-中/低需求的区域(重点保护区),建议发展低管理强度的有机农业,保护区域内的生境;对于中/低供给-高需求的区域(重点提升区、一般提升区),建议使用养殖蜜蜂提升传粉服务,同时在景观尺度上增加自然/半自然生境比例,建立生态廊道以增加生境之间的连通度;对于供需等级较为一致的区域(重点维持区、一般维持区),建议降低农业区域的管理强度,同时增加生境间的连接度,以实现传粉服务的可持续供给。 相似文献
9.
Local habitat structure and resource availability appear to be of great importance for the diversity, abundance, and community structure of bees. We examined the contribution of three different habitat types (farmland, forest edge, forest understory) to bee diversity in a tropical forest-agriculture mosaic in Western Kenya and analysed differences and overlap in plant-bee community interactions between nearby habitats. We used network properties (network size and specialization indices) and bee species turnover to examine temporal and spatial variation of flower-visitation interaction networks in general and in bee species composition in detail across habitats. In total we found 121 bee and 89 plant species involved in the interactions. Results suggest that bees were limited by floral resources because the largest networks, highest diversity, and largest abundances were found at the forest edge and in the farmland which hosted higher amounts of flowers and a more homogeneous distribution of resources in space and time. Forest in the study area is characterized by (1) lower flower density and (2) more humid conditions relative to the farmland. We therefore suggest that the species-rich and structurally diverse farmland acts as a “pollinator rescue” which supports bee communities in the natural forest. We advise conservation managers and politicians to conserve the structural richness of the farmland to (1) preserve bee diversity within the farmland and (2) conserve positive effects of the farmland on bee activity in the forest remnant. 相似文献
10.
Studies of the interspecific relationships between alien and native pollinator species can help forecast the success of alien species as well as assess the extent of disturbance to native plant-pollinator interactions. We examined the habitat and flower resource occupancy by the invasive introduced bumble bee Bombus terrestris and three dominant native bumble bees in central Hokkaido, northern Japan, in relation to a landscape factor (forest cover ratio) and flower morphology with respect to the proboscis length of bees. Three years of monitoring the invasive boundary of B. terrestris indicated that this species, which dominates open agricultural areas, probably will not invade the forests in which the native species dominate. This habitat partitioning likely followed the displacement of the natives by the invader in open agricultural lands. In forested areas, the native species partitioned flower resources on the basis of the relationship between proboscis length and the lengths of the corolla tubes of flowers. However, in open agricultural areas, both the long- and short-tubed flowers were primarily visited by the exotic short-tongued B. terrestris, which foraged illegitimately (by nectar robbing) on long-tubed flowers (Trifolium pratense L.) and legitimately on short-tubed flowers (T. repens L. and Lavandula angustifolia Mill.). The invasion of B. terrestris into open natural vegetation, in addition to open agricultural areas, has recently been reported in Hokkaido. Even though the exotic and dominant native bumble bees partition their habitat according to landscape factors, the invasive bee has the potential to alter the overall interactions within plant-pollinator systems in the regions of open vegetation on Hokkaido Island. 相似文献
11.
Declines in bee populations have been documented in several parts of the world. Bees are dependent upon flowering plants for resources, and flowering plants often depend upon bees for pollination services. Bees can therefore serve as indicator species of habitat degradation due to these relationships with flowering plants. This study investigates how the bumble bee community in San Francisco has responded to urbanization and which urban park characteristics are important for the current community’s structure. To answer these questions we sampled bumble bees, in 18 urban parks and two nearby wild parks. We estimated park characteristics and used multiple regression analysis to determine which characteristics predicted bumble bee abundance and species richness. Bumble bee abundance was positively associated with resource availability or proxies of resource availability; “natural area” (areas that contain remnant fragments that have been largely unchanged by human activity) in 2003 and nest site abundance and openness of the surrounding matrix in both 2003 and 2004. Bumble bee species richness was negatively associated with abundance of a dominant species, Bombus vosnesenskii, in 2004. The importance of the surrounding matrix suggests that these parks do not act as islands. Accordingly, area of park did not explain species richness, while abundance of the dominant competitor did. The species that was most influenced by competition, Bombus sitkensis, uses rodent holes as nest sites and is possibly excluded from nest sites by the early emerging B. vosnesenskii, another subterranean nester. The species least influenced by competition, Bombus melanopygus, is able to use both rodent holes and abandoned bird nests as nest sites. 相似文献
12.
Andreas Müller Stefan Diener Simone Schnyder Katharina Stutz Claudio Sedivy Silvia Dorn 《Biological conservation》2006,130(4):604-615
Knowledge about the quantitative pollen requirements of solitary bees is crucial for the preservation of endangered bee species and the understanding of the evolution of bee–flower relationships. We estimate the number of flowers required to rear a single larva for 41 European bee species (i) by comparing the pollen content of brood cells with the pollen quantity contained in the flowers of the bees’ host plants and (ii) by deducing the pollen requirements from a regression model describing the relationship between the average bee dry body mass and the average brood cell pollen content. The flower requirements of the bee species examined vary by three orders of magnitude. Depending on both bee species and host plant, from seven to 1100 flowers or from 0.9 to 4.5 flower heads are needed to rear a single larva. As only about 40% of the pollen contained in a flower was found to be available to a single female bee, these minimal figures have to be multiplied by a factor of approximately 2.5 to obtain a realistic estimate of bee flower requirements. The amount of pollen lost from flowers for bee nutrition is surprisingly high. We hypothesize that the recent decline of many bee species may have its main cause in a food shortage provoked by a decrease in flower diversity and quantity following habitat destruction and modern agricultural practices. The substantial pollen losses to bees as documented in this study support earlier findings on floral adaptations against excessive pollen harvesting by bees. 相似文献
13.
Organic farming has often been found to provide benefits for biodiversity, but the benefits can depend on the species considered and characteristics of the surrounding landscape. In an intensively farmed area of Northeast Italy we investigated whether isolated organic farms, in a conventionally farmed landscape, provided local benefits for insect pollinators and pollination services. We quantified the relative effects of local management (i.e. the farm system), landscape management (proportion of surrounding uncultivated land) and interactions between them. We compared six organic and six conventional vine fields. The proportion of surrounding uncultivated land was calculated for each site at radii of 200, 500, 1000 and 2000 m. The organic fields did not differ from the conventional in their floral resources or proportion of surrounding uncultivated land. Data were collected on pollinator abundance and species richness, visitation rates to, and pollination of experimental potted plants. None of these factors were significantly affected by the farming system. The abundance of visits to the potted plants in the conventional fields tended to be negatively affected by the proportion of surrounding uncultivated land. The proportion fruit set, weight of seeds per plant and seed weight in conventional and organic fields were all negatively affected by the proportion of surrounding uncultivated land. In vine fields the impact of the surrounding landscape was stronger than the local management. Enhancement of biodiversity through organic farming should not be assumed to be ubiquitous, as potential benefits may be offset by the crop type, organic-management practices and the specific habitat requirements in the surrounding landscape. 相似文献
14.
Secondary vegetation, associated with changes in land use, presents a conservation issue in the preservation of biological diversity in agricultural landscapes. We examine the interactive effect of eucalypt regrowth and rock habitat on reptile species richness and assemblage structure in fragmented agricultural landscapes in south-eastern Australia. Zoogeographic and geomorphic factors influenced species richness and community composition. Saxicolous and arboreal species were less abundant in grassy woodland regrowth, whereas Bassian and fossorial species responded positively to forest regrowth (and tree plantings). Regrowth with rock habitat had higher reptile richness, and more old growth-associated taxa, than regrowth without rock habitat. Thus, the presence of saxicolous habitat can reduce the time required for regrowth to attain a climax community structure and elements of old growth fauna. However, in the absence of vegetation management, secondary vegetation can reduce habitat suitability for a broad range of reptiles. In agricultural landscapes historically supporting savannah-like vegetation, habitat manipulation may be necessary to maintain reptile diversity. Furthermore, tree plantings in temperate agricultural landscapes should aim to restore historical vegetation composition and structure, thereby reflecting the habitat requirements of extant species and facilitating evolutionary processes. In grassy woodland ecosystems, this may involve heterogeneous plantings which emulate natural levels of canopy cover and solar penetration. Maintaining biodiversity in agricultural landscapes will therefore involve managing trade-offs to preserve areas of dense regeneration for regrowth-dependant fauna, while at the same time, creating open-canopy environments to enhance habitat for ectothermic organisms. 相似文献
15.
Pathogen spread or ‘spillover’ can occur when heavily infected, domestic hosts interact with closely-related wildlife populations. Commercially-produced bumble bees used in greenhouse pollination often have higher levels of various pathogens than wild bumble bees. These pathogens may spread to wild bees when commercial bees escape from greenhouses and interact with their wild counterparts at nearby flowers. We examined the prevalence of four pathogens in wild bumble bee populations at locations near and distant to commercial greenhouses in southern Ontario, Canada. Bumble bees collected near commercial greenhouses were more frequently infected by those pathogens capable of being transmitted at flowers (Crithidia bombi and Nosema bombi) than bees collected at sites away from greenhouses. We argue that the spillover of pathogens from commercial to wild bees is the most likely cause of this pattern and we discuss the implications of such spillover for bumble bee conservation. 相似文献
16.
European honey bees used for commercial honey production represent a potential source of competition for floral resources with native nectar and pollen feeding insects. This study reports the results of an experiment run over two years on the impact of commercial honey bees on the fecundity of a solitary native bee, Hylaeus alcyoneus. Registered apiary sites were used as treatment sites (with honey bees) while control sites (without honey bees) were interspersed between. The fecundity of H. alcyoneus was measured using trap nests. We compared the number of nests produced, number of eggs per nest and emerging progeny mass of H. alcyoneus in sites with and without commercial bee hives. The number of nests produced by H. alcyoneus was 23% less (Wilcoxon’s T) at treatment sites than control sites. Analysis of individual measurement intervals using ANOVA was compromised by a general lack of power. This result highlights that even though honey bees have been present in certain areas for many years, competition with native bees may still be occurring. 相似文献
17.
18.
We sampled grasslands at 108 sites, representing five combinations of soil disturbance and enrichment that occurred in eucalypt grassy woodlands in eastern Australia. Intensive land use was represented as a combination of cultivation and fertilizer addition. Other disturbances were represented as various combinations of moderate or no soil disturbance and water enrichment or no enrichment. We identified disturbance response groups that were broadly summarized as Generalists, Tolerant species, Specialists and Intolerant species. The exotic component of the grasslands differs from the native in that there were more Intolerant species amongst natives and more Disturbance Specialist species amongst exotics. However, there were strong similarities, with the majority of exotics (66%) and of natives (60%) being either Generalists or Disturbance Tolerators. The evidence from this study suggests that significant expansion of intensive land use in the district, combining cultivation and fertilization, would pose a major threat to grassland vegetation and would threaten 69 of the 121 native species analysed. 相似文献
19.
David I. King Richard B. Chandler Jeffrey M. Collins Wayne R. Petersen Thomas E. Lautzenheiser 《Biological conservation》2009,142(11):2672-2680
Concern about declines in scrub–shrub bird populations has resulted in efforts to create and maintain habitat for these species. Vegetation within powerline corridors is managed to prevent contact of vegetation with transmission lines, and comprises approximately 2% of all of habitat for scrub–shrub birds in southern New England. Although previous studies have documented the use of powerline corridors by scrub–shrub birds, important questions remain about the factors affecting the quality of corridors as habitat for these species. We surveyed birds and monitored nests on 15 corridors in western Massachusetts during 2002 and 2003 to determine whether scrub–shrub birds occupy and successfully reproduce in powerline corridors, and to identify the principal factors affecting scrub–shrub abundance and nesting success. We found that corridors were occupied by scrub–shrub birds of high regional conservation priority, however, four of seven focal scrub–shrub bird species were scarce or absent in narrow corridors, and the abundance of these species was highest in corridors of intermediate width. Overall, nest survival was low (0.14) at these sites relative to other types of early successional habitats in the region, however, if we consider only our sites that were wider than the median width (49 m), nest survival in corridors was (0.33), similar to survival rates reported in other studies of scrub–shrub birds. We conclude that powerline corridors provide habitat for early successional birds of conservation concern, with wider corridors (50 m) contributing more to regional conservation of these species. 相似文献
20.
The effects of grazing on the richness of understorey plant communities are predicted to vary along gradients of resources and tree cover. In temperate Australia livestock management has involved phosphorus addition and tree removal but little research has examined how the effects of grazing on plant species richness may vary with these management regimes. Patterns of understorey plant species richness were examined in 519, 0.09 ha quadrats in grazed pastures and remnant grassy forests and woodlands in southern Australia. Sheep grazing was the primary land use and sites varied widely in grazing frequency and density, tree cover and phosphorus fertiliser history. Using an information theoretic approach the available data provides strong evidence that the effect of grazing on total species richness varies according to available phosphorus and tree cover. Intermittent grazing and no grazing were associated with high total and native plant richness, but only at low phosphorus concentrations. Phosphorus was strongly negatively correlated with richness, particularly at low grazing frequency. Total species richness was positively correlated with tree cover except under frequent grazing at high stocking rates, suggesting that heavy grazing eliminates spatial and temporal heterogeneity imposed by trees. Native plant species richness was negatively correlated with a history of cultivation, positively correlated with tree cover and varied according to landscape position and geological substrate. Frequent high density grazing, particularly when associated with clearing, cultivation and fertiliser addition, was associated with the persistence of very few native plant species. In contrast, the richness of exotic plant species was relatively invariant and performance of the best model was low. While several studies have highlighted the importance of the grazed and cleared matrix for the conservation of native plant species, this benefit may be limited in landscapes where intensive grazing management systems dominate. Strong evidence for interactions between grazing, phosphorus and tree cover suggest that failure to consider other land use practices associated with grazing management systems could lead to erroneous conclusions regarding vegetation responses to livestock grazing. 相似文献