中文摘要：

以寿光蔬菜产业集群为例,运用Arc GIS 10.0软件的空间邻近分析,Ucinet软件的社会网络分析以及多元线性回归模型从多维邻近性视角探究蔬菜集群企业的地理邻近、关系邻近特征及两者在集群企业创新中的影响。研究表明：1蔬菜集群企业具有明显的地理集聚特征,企业的空间邻近有利于集群创新氛围的形成;2集群企业及各创新主体间拥有紧密的关系邻近网络,关系网络中多核心节点,促使创新资源、隐性知识、新技术等的扩散与传播;3进一步回归分析表明,关系邻近对蔬菜产业集群创新作用更显著,永久性地理邻近在寿光蔬菜产业集群创新中仍然起到正向作用,但作用要弱于关系邻近。说明多元关系邻近在集群创新过程中能够对消除过度的地理邻近、单一关系邻近造成的创新锁定起到作用,同时还为全球化背景下农业集群创新提供新的渠道。

英文摘要：

The article aims to analyze the impact of geographical proximity and relational proximity of innovative actors on the innovation of agricultural clusters. Taking Shouguang vegetable industrial cluster in Shandong Province, China as a case study, it examines how the geographical proximity and relational proximity among heterogeneous enterprises influence the innovation effects of the cluster, combining the spatial proximity analysis with the social network analysis and the multiple linear regression model, based on the fieldwork and questionnaire data. It is shown that： 1） The spatial agglomeration of enterprises in cluster is conducive to the formation of innovation atmosphere. The leading enterprises have a high degree of geographical proximity with the new agricultural enterprises, while the processing agricultural enterprises distribute relatively decentralized. Farmers collocated primitively evolve into agricultural enterprises through organizational innovation,which results to enterprise agglomeration, division of labor and close connection with each other, then the cluster forms; At the same time, the temporary geographical proximity among enterprises in the cluster, which is the spatial distance between the enterprise and the local vegetable expo park, mirrors the participant performance of enterprises to the meeting network. 2） The enterprises in cluster have close social connection to other innovation actors, and many core nodes in relational network take crucial roles to the allocation of the innovation resources and the diffusion of the tacit knowledge and new technology not only inside cluster but outside cluster as well. The concentration degree of the relational network of enterprises within cluster is lower than that of the whole network, which indicates that the concentration degree of the cluster＇s external nodes is higher, and the geographical proximity of enterprises is not consistent with their relational proximity to the cluster enterprises. 3） The regression analysis presents that,