The research characteristics
A total of 837 publications on legume mycorrhiza had been retrieved in WOS core collection from 1985 to 2020. The first article was published in PHYTOCHEMISTRY in 2005 in this field. The annual change of publication in the field was presented in Fig 1, which indicated that publications in the field increased markedly with the change of years from 2009 to 2020 (R
2=0.5196). There were not any studies in the research of legume mycorrhiza between 2006 and 2008.
Country collaboration maps showed visually that the collaboration research on legume mycorrhiza involved in 78 countries (Fig 2). Their cooperation characteristics in top 20 countries suggested that the 5 most active collaboration countries are USA, France, Germany, China and Australia (Table 1). However, the highest centrality of countries produced changes with the top 5 countries of Australia, France, USA, India and China, which indicated that these 5 countries contributed more among international exchange in the field than other countries (Fig 2, Table 1).
The network of institution collaboration showed that 931 institutions collaborated with each other to publish papers in this field (Fig 3). Among them, University of Western Australia and Chinese Academy of Sciences contributed a greater number collaboration than other institutions with 33 and 30 times, respectively. Further, the institutions with strongest citation bursts suggested that there were four institutions bursts with the value from 3.7119 to 3.9420 in the field (Table 2).
The annual publications are usually considered as an important indicator for reflecting development and evolution in given research field (
Mongeon and Paul-Hus, 2016). The current result showed that the study in legume mycorrhiza was paid more and more attention (Table 1), which can be expected to be more in-depth and to be improved in this field. Collaboration maps among countries and institutions drawn by CiteSpace is able to identify the relationships among scholars, countries or institutions in a given research field, which provide the basis for evaluating their academic influence and discover noteworthy countries and institutions
(Xing et al., 2020). France is more willing to international exchanges and cooperation recently with 2 institutions in strongest citation bursts on the study of legume mycorrhiza
(Aleixandre-Benavent et al., 2017).
Intellectual structure analysis
Five key document co-citation clusters are obtained (Table 3) based on co-citation constructed network with the modularity Q value of 0.6559 and mean silhouette value of 0.2833 (Fig 4). Further, the top ten cited literatures were listed in Table 4. The top ranked item by citation counts is Maillet F (2011) in Cluster #1 sourced from Nature with citation counts of 95.
Clusters are considered well structure in CiteSpace when modularity is more than 0.3. Larger silhouette values indicate higher homogenization of nodes in the cluster and “silhouette” >0.7 generally suggests that the cluster has high credibility (
Chen, 2017,
2018). In this study, the modularity Q value is 0.6559, which indicates the clustering effect is significant. However, the low mean silhouette value of 0.2833 is probably caused by numerous small clusters
(Yan et al., 2020). The bigger burst value means the more rapid growth of citation frequency
(Yang et al., 2019).
Research trend
Eleven clusters were obtained based on keyword clustering analysis (Fig 5). The clustering characteristics of 4 largest clusters are summarized in Table 5. The modularity Q and mean silhouette value of keyword clustering network are 0.4836 and 0.4901, respectively. The largest cluster (#0) has 79 members with silhouette value of 0.757 labeled as
Lotus japonicus.
A total of 25 burst keywords are identified in the keyword co-occurrence network. Twenty-four keywords with high burstiness were identified and are listed in Table 6. These keywords show the research topics in the research of legume mycorrhiza and their changes in different periods. Begin and end year following each keyword indicates the duration of citation bursts. The top ranked of keyword by bursts was rhizobia with the bursts strength value of 5.2899, which began from 2016 and ended in 2018. Further, assessing the 24 keywords according to their bursting period, it shows that similar studies were able to be paid repeatedly attention in different periods, such as “arbuscular mycorrhizae” and “arbuscular mycorrhiza fungi”. Meanwhile, the distinction of some keywords is relatively small, such as “nod factor”, “root nodule” and “rhizobia”.
In the constructed network of keywords, the Q value of the modularity was more than 0.3. Therefore, the clustering effect was highly significant
(Li et al., 2020). The clustering network of keywords is almost reasonable with mean silhouette value of 0.4901 because a cluster with a silhouette >0.5 is considered reasonable (
Chen, 2004;
Guo et al., 2020). In the CiteSpace co-occurrence analysis, burst keywords usually reflect topics that have attracted the attention of peer scientists and are often used to explore the hotspots and research frontiers of a research field
(Zhang et al., 2019, 2020). Therefore, the research hotpot and trend should be interaction between legume plants and mycorrhizal fungi for improving nutrition absorption, N-fixation, resistance to stress and their mechanisms.
The visual bibliometric analysis of this study bridged thoroughly the gap between extant literatures in the study of legume research and their characteristics including changes with years, publishing institutions and their cooperation,
etc. because much attention has been given to legume research. However, no previous studies mapped out the linkage or working relationships among the clusters of research institutes and their countries in this field.
Previous studies have never analyzed deeply research corpus including such broad aspects such as collaboration of country or institutions, co-citation clusters, cited references, keywords and all kinds of visual maps.
The most obvious advantage of CiteSpace is that it allows scholars to investigate the specific research areas by analyzing the citations, co-citations, keyword burst and cluster, thus drawing very useful visual conclusion. At the same time, the analysis of the network of cooperation among countries or institution, document co-citation and keyword clusters by CiteSpace based on published literatures in the WOS core collection are helpful for obtaining more accurate and complete information in the international study in this field. Further, quantified analysis, intellectual structure and emerging trends provided insights into research hotpots and trend by different perspectives in the future.
The significance of this study provided an overall picture and development path for legume mycorrhizal research to help scholars to broaden their research horizons and determine future research directions. The overall analysis of research characteristics also provided an important reference for academic research and relevant policy formulation in the future.
However, there were some limitations in this current study. For examples, I did not collect all the related literatures published in non-english language. Some similar keywords are not able to be combined in the software of CiteSpace. In my opinion, future literature mining should consider more databases as well as non-english literatures and blending analysis of similar keywords based on these limitations. Further, I also suggest conducting more in-depth content interpretation and analysis in future studies.
