3. Result
3.1 Basic characteristics of the literature
3.1.1 Quantity of articles and citations
From 1985 to 2020, the number of articles published in the field of seed dispersal worldwide increased, indicating that this field has received extensive attention from scholars and has strong developmental potential (Fig. 2). From the perspective of the number of articles, there were two turning points in the publication of seed dispersal research. One turning point was in 1993, when the number of articles published exceeded 31 for the first time; prior to 1993, the number of articles published annually grew very slowly. The other turning point was in 2002, when the number of articles published exceeded 47 for the first time. Since 2002, the number of articles published has increased steadily every year. From 2002 to 2020, an average of 123.33 articles were published each year, with an average annual growth rate of 4.21%. This result indicates that research on seed dispersal is in the growth stage and that this field has very large research potential. During the period from 1985-2020, the total number of citations was 72,460; the number of citations per article in this period was 37.49. The average number of citations was highest in 2000, reaching 146.63. Subsequently, the average number of citations showed a downward trend.
3.1.2 Analysis of subject categories
The 10 most frequent subject categories according to our research are shown in Table 1; these subjects include ecology (1,199 articles, accounting for 61.84% of the total), plant science (424 articles, 21.87%), evolutionary biology (204 articles, 10.52%), forestry (168 articles, 8.66%), biodiversity conservation (159 articles, 8.20%), zoology (141 articles, 7.27%), biology (101 articles, 5.21%), multidisciplinary sciences (99 articles, 5.11%), environmental sciences (89 articles, 4.59%), and genetics and heredity (86 articles, 4.44%). The number of publications in each category reflects the trends of seed dispersal research in different domains. As shown in Table 1, we found that the number of publications in the fields of ecology and plant science accounted for the vast majority of the literature included in this analysis.
Fig. 3 displays a co-occurring subject category network that includes 62 nodes and 227 links. Through the subject category network, we found that seed dispersal research is a discipline that involves various subjects, such as ecology, plant science, forestry, zoology and biology. Fig. 3 shows that the 3 most frequent categories in seed dispersal research are ecology, environmental science and plant science. Environmental science has high centrality because it connects many different subject categories, including atmospheric sciences, ecology, environmental chemistry and geoscience. Ecology also connected many subject categories, including environmental science, plant science, forestry and biodiversity conservation.
3.1.3 Journal analysis
The 1,939 articles on seed dispersal analysed in this study appeared in 371 journals. Table 2 lists the 10 journals with the highest numbers of publications on seed dispersal. The number of articles published by these 10 journals in the field of seed dispersal encompassed 27.7% of all the analysed publications. Biotropica was the most popular journal (88), followed by Oecologia (66), Journal of Tropical Ecology (65), Journal of Ecology (62) and Ecology (61). The number of total citations of a journal (TC) and the average number of citations per paper of a journal (TC/P) reflect the impact of that journal (Ji et al. 2014). The impact factor (IF) and H-index can also measure journal impact. As shown in Table 2, among the literature included in this study, Biotropica, Oecologia, Journal of Tropical Ecology, Journal of Ecology and Ecology each published more than 60 articles about seed dispersal. Ecology had the highest impact in the field of seed dispersal, with the highest H-index (41), TC (5614) and TC/P (92.03).
3.2 Research power of seed dispersal
3.2.1 Analysis of countries and institutions
(1) Quantity of articles and citations
The papers included in this study came from 82 countries around the world. The 10 most productive countries are listed in Table 3. As shown in Table 3, the USA had the largest number of publications (555 articles) and total citations (26,942 citations). Spain had the second highest number of publications (214 articles), followed by Brazil (156 articles), Germany (149 articles), Australia (109 articles) and China (96 articles). Fig. 4 shows the distribution of countries/territories and the relationships of international cooperation with impact on seed dispersal research. The USA had the earliest publication on seed dispersal and was the most productive country during the study period. The USA had the highest degree of centrality (0.60), followed by France (0.21), Germany (0.20) and Australia (0.16); these countries play an important role in collaboration networks. The analysis revealed that there was strong international cooperation between North America and Europe; however, China’s international cooperation was weak.
(2) Collaboration network
Table 4 lists the 10 most productive institutions in terms of relevant articles. Consejo Superior de Investigaciones Científicas (CSIC) was the most productive institution, followed by the Chinese Academy of Sciences, Universidad Nacional Autonoma de Mexico, the University of Florida and Kyoto University. Fig. 5 shows the network of academic collaboration on seed dispersal research between institutions. The strongest collaborations were identified between CSIC and the University of Oviedo. Kyoto University collaborated with the Forestry & Forest Product Research Institute. The University of Oviedo had the highest centrality (0.16) and cooperated closely with other institutions, such as CSIC and the University of Sao Paulo.
(3) The development of seed dispersal research in selected countries
To evaluate the changes in the academic impact of the 10 abovementioned countries in terms of seed dispersal research, the AI and the AAI were used. It should be noted that because there is usually a lag between the publication time and the citation time of an article(Glänzel and Persson 2003, Qiu and Chen 2009), the time range of the AAI was set 2 years behind that of the AI in this study.
Quadrant diagrams of the changes in the two indexes are shown in Fig. 2, with quadrants I to IV representing 4 different situations. The points in quadrant I represent the years in which the country’s AI and AAI indexes were both higher than the global average levels; the points in quadrant II represent the years in which the country’s AAI and AI indexes were higher and lower than the global average levels, respectively; the points in quadrant III represent the years in which the country’s AI and AAI indexes were both lower than the global average levels; and the points in quadrant IV represent the years in which the country’s AI and AAI indexes were higher and lower than the global average levels, respectively.
In general, with the exception of Australia, the AI and AAI indexes of the 10 selected countries showed an upward trend during the study period. As shown in Fig. 2, the research power and academic impact of the USA in most years were higher than the global average level; in contrast, although Spain had a relatively large number of published and cited articles, its academic impact was lower than the global average level in most years. The research power and academic impacts of Brazil, China and Japan were lower than the global average level in most years, but in recent years, their academic impact was higher than the global average level, indicating that the impact of these countries on seed dispersal research is continuously increasing.
3.2.2 Author analysis
(1) Author collaboration network
Table 5 lists the 10 most productive authors according to the publications included in this study. As shown in Table 5, A Traveset, M Nogales and P Jordano had the highest numbers of publications. These authors were followed by XF Yi (19 articles), M Galetti (19 articles), D Garcia (18 articles), R Nathan (17 articles), SB Vander Wall (15 articles), MA Pizo (15 articles) and MB Soons (13 articles). The 10 most productive authors all came from different research institutions (Table 5); 4 of these authors were in Spain, and the remaining 6 were in different countries.
Fig. 7 is the collaboration network map of the authors of seed dispersal research. Five author groups were identified. (1) A team consisting of A Traveset, P Jordano and M Nogales (circle A) focused their research on long-distance seed dispersal (Nogales et al. 2012, Traveset et al. 2014, Pedro and Jordano 2017). (2) A group of authors led by XF Yi (circle B) focused on animal-seed interactions (Steele and Yi 2020, Xianfeng et al. 2020). (3) M Huynen and A Albert formed a core team (circle C), and much of their research focused on the analysis of feeding ecology and seed dispersal of pigtail macaques (Macaca nemestrina ) (Latinne et al. 2008, Aurelie et al. 2013). (4) A team consisting of R Heleno and S Timoteo (circle D) studied ecological networks (such as food webs and pollinator networks) (Timóteo et al. 2018, Heleno et al. 2020). (5) A group of authors led by M Schleuning (circle E) focused on plant-animal interactions and ecological networks (Schleuning et al. 2015, Schleuning et al. 2016).
(2) Author co-citation network
The results of the author collaboration analysis can reflect an author’s contributions to the field of seed dispersal and the cooperative relationship between authors, but it cannot reflect an author’s impact on seed dispersal research. Therefore, we used author co-citation analysis to provide further information. We selected the 10 most cited project themes from each topic to ensure that we targeted the most prominent authors.
On the basis of the author co-citation network analysis results, the authors with high citation frequency and their countries of origin were determined. As shown in Table 6, the author with the highest citation frequency is R Nathan (3,612 citations, Israel), followed by P Jordano (2798 citations, Spain), DJ Levey (1273 citations, USA), A Traveset (1055 citations, Spain), M Galetti (976 citations, Italy), MB Soons (933 citations, Netherlands), SB Vander Wall (853 citations, USA), D Garcia (832 citations, Spain), M Nogales (637 citations, Spain) and PR Guimaraes (634 citations, Brazil). Four of these authors are from Spain, two are from the USA, and the others are from Israel, Italy and the Netherlands.
By comparing Table 6 and Table 5, we determined that there is a strong relationship between the 10 most cited authors and the 10 most productive authors. Only DJ Levey, M Nogales and PR Guimaraes appear in only Table S2; the remaining authors appear in both tables.
3.3 Intellectual base
The number of citations of an article represents the degree of recognition it received in a particular research field or its academic impact (Ebrahim et al. 2013, Yoshikane et al. 2013). The 10 most highly cited references are listed in Table 7. The most highly cited reference was published by R Nathan, with a citation count of 1,343. Through this review, the author shows that with the growing interest in spatial ecology, new methods for researching seed dispersal, which is one of the key processes in determining the spatial structure of plant populations, are emerging. Seed dispersal methods vary among plant species, populations and individuals; with distance from the parent plant; and by microsites and times. The development and improvement of mathematical models is expected to produce a deeper and more mechanical understanding of the seed dispersal process and its consequences. The remaining documents with a high citation rate were published in the 21st century, mostly between 2000 and 2011. The content of these papers shows that the intellectual base focuses on long-distance seed dispersal and dispersal patterns in different habitat conditions. These publications have laid a good foundation for future research directions.
3.4 Research hotpots
3.4.1 Keyword co-occurrence
Keywords reflect an author’s intention and interests and summarize the main contents of an article. People can obtain useful information, such as research goals, methods, and important opinions, from the keywords of an article (Xu et al. 2018). Therefore, keyword frequency analysis and period change analysis are essential to discussing research hotspots and development changes within a research field (Wang et al. 2018).
We use the keyword co-occurrence analysis in CiteSpace software to create a keyword co-occurrence network diagram of the seed dispersal research analysed and identify the main research hotspots in this field (Fig. 8). Each node in the graph represents a keyword; larger nodes indicate keywords that appear more frequently. Similarly, the width of the links indicates the frequency of keywords appearing together; a thicker link indicates that the two connected keywords appear together more frequently. A total of 624 keywords were obtained, among which 14 appeared more than 100 times. There is no doubt that “seed dispersal”, the main keyword of this article, is the largest node in the Fig. 8; this keyword appears 324 times and is closely related to “pattern” (245 times), “plant” (165 times), “frugivory” (160 times), “recruitment” (160 times), “ecology” (152 times) and “forest” (152 times). Other keywords, such as “germination” (136 times), “bird” (122 times), “predation” (136 times), “rainforest” (101 times) and “distance” (100 times), also had high frequencies.
3.4.2 Keyword burst analysis
Keywords burst analysis identifies those keywords with a sharp increase in frequency. Burst detection is a useful analytical method for finding keywords that are of particular interest to the relevant scientific community over a given period of time (B et al. 2018). Therefore, bursting keywords can be used as indicators of research frontiers and to forecast research trends. Fig. 9 shows the 20 keywords with the strongest citation bursts from 1985 to 2020. In this figure, the strength of the burst represents the intensity of focus on a given topic. In the field of seed dispersal research, wind dispersal has been a research hotspot since 1989, with the focus subsequently diversifying to include “selection” (strength: 4.55), “bird” (strength: 5.56), “dispersal” (strength: 5.26), “environment” (strength: 5.21), “tropical rainforest” (strength: 8.2) and “dynamics” (strength: 5.01). We found that in the last ten years (2010-2020), studies related to “conservation” and “spatial pattern” were emerging active topics (Fig. 9).
3.4.1 Keywords in different countries
Due to differences in climate, geographical features, and historical and economic conditions, the development of research in a certain field in different countries or regions may not be balanced. Fig. 10 shows the high-frequency keywords, main research types and regional distribution of relevant research in the 10 countries with the highest numbers of publications. The results showed that in the field of seed dispersal, the differences between countries or territories are not very large and that the study of the relationship between ”seed dispersal” and ”frugivorous animal” received the most attention. For example, from 1985 to 2020, the most frequently used keywords in the USA were “seed dispersal” (162 times) and “frugivory” (75 times); the most frequently used keywords in Spain during this period were “seed dispersal” (41 times) and “frugivory” (26 times).