3.4 Keyword co-occurrence analysis
Research topics and hot spots in a period can be obtained through co-occurrence analysis of keywords, and research frontier of three periods can be analysis accordingly (Ouyang et al., 2018; Yu et al., 2017). As shown in Fig. 4, 5 and 6, with time going on, the nodes of keyword number are increasing and diversified, and the density of the keyword co-occurrence is also getting stronger. In addition, the first 15 keywords related to hydrological connectivity in the three stages are also listed in Table 3. The analysis of keywords is discussed based on frequency and centrality according to Fig. 4 to 7 and Table 3.
==============================================
Please Place Table 3 and Figures 4 to 7 here
==============================================
Fig.4 shows the co-occurrence network of keywords during 1998-2004. It can be seen that the keyword “ecosystem” is the main research content in this period and has high co-occurrence with connectivity, wetland, floodplain and biodiversity. During this period, land use change and river reconstruction in European countries lead to a serious decline in the hydrological connectivity of floodplain wetland which further affects the ecosystem and ecological functions (Gumiero et al., 2013). As a sequence, with the decrease of hydrological connectivity of floodplain, the richness of species and diversity of ecosystem gradually decrease. According to the Flood Pulse Concept (FPC), the exchanges of nutrients and sediments between river and its floodplain is a more significant ecosystem function than those of longitudinal processes (Aarts, Van den Brink, and Nienhuis, 2003). Hence, scholars in this period have done a lot of research on the effects of floodplain on ecosystem and biodiversity. Chovanec and Waringer (2001) believe that horizontal connectivity is an important feature of ecological integrity in floodplains, which is coincided with the main idea of FPC. By analyzing the breeding behavior of dragonflies, he provided a new method for the ecological integrity evaluation of the river-floodplain system. In addition, although hydrological connectivity is the theme of this paper, it did not appear in the top 15 of the keyword co-occurrence networks at this period, which indicates that hydrological connectivity had not been the major research focus at this period.
With the passage of time, the number of nodes in the keyword co-occurrence network is increasing, and the connection density between each node is significantly enhanced, which means that hydrological connectivity attracts increasing attention in this period (Fig. 5). As shown in Table 3, during the period from 2005 to 2011, the ecosystem maintained a high co-occurrence frequency and centrality, which was inseparable from the hydrological connectivity and the close connection between the ecosystem and biodiversity. The word “hydrological connectivity” appeared and rank second in the top 15, indicating that the research topic of hydrological connectivity developed rapidly at this period. Furthermore, the “water” and “hydrology” rose up to the top five, while “wetland” “floodplain” and “biodiversity” have heated down or even drop out of the top 15, this might because that the researchers in the previous period were mainly from the field of ecology, and more researchers doing hydrology and water resources were involved in 2005-2011. In addition, the frequency of node “dynamics” has increased significantly, and its centrality suggests it has attracted more attention in this period. The hydrological connectivity of water system has a decisive influence on the hydrological dynamic characteristics. With the intensification of human activities at this period and the improvement of urbanization, the utilization measures of rivers, such as the construction of dykes and dams, changed the hydrological dynamic conditions of rivers, thus blocking the hydrological connectivity of rivers.
The keyword co-occurrence network from 2012 to 2018 is shown in Fig. 6. It can be seen, in recent years, the study of hydrological connectivity has made great progress. “Ecosystem” still occupies a major position as a key point of hydrological connectivity. In addition, “climate change” first appeared in top 5, which means the impact of climate change on hydrological connectivity has gradually entered the research focus of many scholars. In the past half century, as the climate warms, the total amount of global river runoff tends to decrease, the frequency and intensity of extreme hydrological events increase, and the hydrological pattern of wetlands has undergone profound evolution and developed in the direction of severe decrease of hydrological connectivity. Future climate projections for the Amazon indicated a rise of temperature (Ipcc et al., 2013), total rainfall is likely to decrease (Malhi et al., 2008). Such dry-warm conditions would likely reduce the capacity of ecosystems to store water and buffer emissions and increase the frequency of low-water events in large rivers (Costa, Botta & Cardille, 2002), lead to the interruption of hydrological connectivity, which adversely affects the local ecosystem, hydrological dynamic processes and water quality accordingly.
Though “management” does not have a high frequency as “climate change”, the number of nodes co-occurrence with “management” are more than “climate change”, and the centrality is higher, too. Facing the problems of water resources, environment and ecology brought by climate change and human activities, it is a good management method to build hydrological connectivity projects of rivers, lakes and marshes to face the needs of water supply, flood control and drought prevention, as well as water quality and ecological security. Hence, in recent years, scholars have conducted a large number of studies on the characteristics, classification systems and influencing factors of river and lake system connectivity, aiming to provide technical support for the construction planning and operation management of hydrological connectivity projects. Therefore, the topic related to hydrological connectivity at this period mainly focus on the comprehensive impacts of climate change and human activities, including but not limited to water quality, hydrological hydrodynamics and water ecosystem.
3.5 Author co-cited analysis
The network of co-cited authors is shown in Fig. 7. Compared with the total number of articles published by the author, co-citation analysis is more representative of the value of published articles. It can be seen in Fig. 7 that Ward has the most co-cited frequency, account for 461, followed by Junk (409) and Tockner (396), and the top three co-cited authors also have strong centrality, which means their publications were accepted by more scholars. A citation burst means that a reference has a sharp increase in citations in one year or multiple years, which can reflect whether a publication is hot or not for the research area (Yu, 2015). Table 4 lists the top 15 co-cited authors with the Strongest Citation Bursts. The top ranked is Bracken L J with bursts strength of 30.3936, second one is Tetzlaff D, 21.2137, followed by Schiemer F (17.6171). In addition, we found that the burst detection red line of some co-cited authors expired several years ago (Schiemer F), which means that he had a strong burst period and was the hot topic at that time, but now the heat has decreased. While some authors’ (e.g., Bracken and Tetzlaff) burst detection red lines extend to 2018, indicating that the author’s articles still have high hot spots in recent years which may continue in the future.