Full bibliometric data of this set of documents were exported in the form of text files for analysis, which included their title, date of publication, author names and affiliations, citation count, list of keywords, abstract text and list of references. The source of literature data is Web of Science database core collection (accessed on 9th to 22nd September 2023),¹ which the retrieval period was from 2000 to 2022. To analyze the grass-livestock interaction research in its broadest scope, collective publications in grass-livestock interaction were considered and their full bibliometric information was exported and analyzed. The specific research steps: (1) Web of Science core database searching subject is grass and livestock with a total of 18,342 valid data; (2) control time range is from 2000 to 2022, with a total of 11,636 data; (3) the type of control literature is articles and reviews, with a total of 8,258 data; (4) 5505 valid literature records are obtained after refining, cleaning, and summarizing the initial search results to ensure the quality and applicability of literature data.

The research methods of this paper include bibliometric method, content analysis method and clustering analysis method. In this study, the methodology of Visualization of Similarities (VOS) and the Document Co-Citation Analysis (DCA) are proposed to process the retrieved literature data, and data mining, mapping, and clustering are carried out graphically to extract key information from countries, academic institutions, journals, and keywords. They can be constructed based on citations, bibliographic coupling, co-citations, or co-author relationships. Compared with other document metrology software, VOSviewer has the advantages of strong graphical display ability, which is suitable for large-scale data, and strong versatility, which is suitable for various formats of source data. The clustering algorithm employed in VOSviewer is the stepwise clustering approach, which involves R-type clustering analysis of a large samples (keywords) and their categorical variables (such as citation frequency, publication year, etc.) after filtering, and primarily relies on the algorithm of association strength, selecting high-frequency keywords from the study for cluster analysis (van Eck and Waltman, 2017). The analysis results include cluster divisions, keyword linkage coefficients (co-occurrence strength), total node linkage strength (reflecting the strength of relationships between multivariate nodes), and the strength assignment of various variables related to keywords.

The keyword clustering (Table 3) and visual knowledge map formed by the selected relevant literature data are shown in Figure 7. Keywords related to the grass-livestock interaction can be divided into four clusters. The cluster one is the cluster related to “Livestock and grassland forage” composed of the red area, which mainly involves the contents related to animal husbandry such as cattle, sheep, grass, grazing, grassland, pasture, feed quality, feed yield, feed nutritional value and so on. It contains 2,389 keywords. Among them, Cattle, Nitrogen, Growth, Grass, Sheep, Grazing, Grassland, Performance, Pasture, Quality, Soil, Forage, Digestibility, are high-frequency keywords that are widely cross-linked with keywords in other clusters. The cluster two is target “Vegetation and grassland management” for the green area, which includes management strategies, biomass change, species richness, and sustainability. It is composed of the keywords “Management, Vegetation, Dynamics, Systems, Biomass” and so on. The cluster contains 1,336 keywords. The keyword “management” has the highest frequency (633 times) and the highest total link strength (6,434 times). Meanwhile, the top five terms (Cultural Landscape, Semi-Natural Grasslands, Sustainable Grazing, Pika Ochotona-Curzoniae, Temperature Sensitivity) with the highest standardized average citation rate are all from this cluster. The cluster three is related to “Environmental impact and resource conservation” for the blue region containing 1,176 keywords, which mainly deals with the role of animal husbandry in the restoration and protection of grassland ecosystem, and the relationship between animal behavior and ecology. It is composed of the key words “Livestock, Diversity, Biodiversity, Productivity, Conservation” with high repetition frequency. The average number of citations is the highest, which indirectly indicates that the keywords related to “Livestock” have received early attention and continue to affect livestock balance. The cluster four is composed of “Forage characteristics and soil texture” clusters in the yellow regions containing 1,065 keywords, which deals with different kinds of grass herbs and their relationship to soil phosphorus, nutrients, bacteria. Top keywords are Grass, Plant, Phosphorus, Tall fescue, Forage quality, Ryegrass, White clover, Cultivars, Environment. As the basis of forage feed, the keywords related to soil ecological environment play important roles in the study of grass-livestock interaction.

The initial exploration period (2000–2011) is the initial stage of the Millennium Development Goals (MDGs) put forward by the United Nations. Animal husbandry is the pillar industry in many developing countries, and researches in the field of grass-livestock interaction have attracted wide attention. However, researches on grass-livestock interaction in countries around the world tend to adopt a single-factor and crude analysis model at this period. In addition, the direct factors that threaten animal husbandry and anti-risk strategies have been explored worldwide to cope with climate change and food crisis (Li, 2009). The research on grass-livestock interaction in China is in the initial stage of development, and the research content is mainly to establish a regional water-grass-animal mathematical model according to the planned area (Jun Li et al., 2007; Zhou et al., 2017), propose the theoretical threshold of grassland ecosystem management (Li et al., 2005).

During the period of rapid development (2012–2016), developing countries, which have long been plagued by problems such as pasture shortage, poverty (Briske et al., 2015), and food security, intend to achieve increased production and income through research related to grass-livestock interaction in order to accelerate the realization of the United Nations Millennium Development Goals (Yang et al., 2019). At this stage, the research goal of grass-livestock interaction was inclined to promote economic growth and social equity (Wang et al., 2017; Zhang and Wang, 2023). By referring to the keywords, it is found that the research of Chinese scholars is consistent with the policy direction. During this period, China introduced a series of protection policies, including grassland protection and construction, economic development of pastoral areas, and grassland supervision system (Li et al., 2018, 2021). At the same time, developed countries pay attention to the balance of ecological protection and biomass, focusing on the role of human activities in the grass-livestock interaction (He et al., 2015; Li et al., 2017; Wang et al., 2017; Wang S, et al., 2023), or the study of global grassland degradation or grassland ecosystems under extreme conditions (Yao et al., 2016; Dong et al., 2020; Bardgett et al., 2021).

In the transitional development period (2017–2022), policies and research in countries around the world begin to adapt to new challenges, and the grass-livestock interaction field integrates climatology, ecology, sociology, etc., to deal with more complex problems (Sun et al., 2020). Furthermore, all-round and full-detailed research on professional fields emerged (Rosolem et al., 2017). The research on grass-livestock interaction shows the trend of comprehensive and sustainable development emphasizing the economic, social, and ecological protection.

In the grass-livestock interaction field, the number of publications has shown a significant increasing trend in the recent 23 years. The number of keywords related to “Management, Livestock, Vegetation” has increased (Figures 4–6), covering multiple disciplines such as botany, soil science (Pulido et al., 2018), animal management (Reinermann et al., 2020), climate change (Uwizeye et al., 2020), social economy (Sharma et al., 2021), and policy science (Wei et al., 2021), which influence social stability, economic growth, and environmental protection (Herrero et al., 2013). Therefore, cross-disciplinary and cross-field cooperation is urgently needed to integrate expertise and research methods in different fields in order to solve scientific problems in the grass-livestock interaction field.

Through summarizing the past research (Figure 2C), it was found that in the initial exploration period (2000–2011), China implemented the subsidy and incentive policy for grassland ecological protection, and the grassland policy was changed from “taking grain as the key link and comprehensive development” to “attaching equal importance to ecological production and giving priority to ecology” (Li et al., 2022), so as to further improve the policy system related to grassland ecological restoration such as returning grazing to grassland, banning grazing and ending grazing. Compared with other countries (Figure 2C), French animal husbandry is faced with short-term or long-term drought at this stage, and how to enhance the natural resilience of grassland ecosystem to cope with various problems caused by climate change is their primary issue (Rigolot et al., 2014). In developing countries with more natural grassland resources, such as Kazakhstan and Mongolia, there are relatively few reports on the research and practice of restoration of degraded grasslands, while in South America, Australia, Africa and other savanna areas, have been partially transformed into pasture, farmland and forestland due to local natural and economic conditions, which the degradation of grassland ecological resources has not attracted enough attention (Gao and Ding, 2022).

During the period of rapid development (2012–2016), the research priorities of developed and developing countries differ greatly (Figure 2C; Distel, 2016; Fynn et al., 2017; Rosolem et al., 2017). Fensham et al. (2015) suggests that land clearing exacerbates the spread of buffel grass and the control of this practice is an important contribution to the conservation of savannas. Scasta et al. (2016) proved that overgrazing, fire, and species invasion greatly damaged local ecosystems in different regions in the United States. In developing countries, faced with problems such as poverty and food shortage, the research direction is mostly to solve social practical problems (Rudel et al., 2015). Distel (2016) evaluated the ecosystem and economic situation under different management modes of pastures in the semi-arid region of central Argentina. Forrest et al. (2016) large-scale removal of landscape-bearing woody plants was urgently needed to promote livestock productivity.

In the transitional development period (2017–2022), the research on grassland productivity (Petrie et al., 2018; Su et al., 2022), ecological protection (Hu et al., 2021), forage management (Guuroh et al., 2018; Arimitsu et al., 2021; Domiciano et al., 2021), anti-risk countermeasures (Fernández-Rodríguez et al., 2018; Davies et al., 2022), and economic benefits (Bilancia et al., 2020; Euclides et al., 2022) related to grassland and livestock interaction has a certain basis (Figures 4–6), and the research method based on interdisciplinary optimization has provided new perspective and brought innovative development to the field of grassland and livestock interaction. In the previous study, Sun et al. (2020) considered grass-livestock-human system as a coupled social ecosystem, and used the theories and methods related to population, resources and environmental economics to explore strategies on the development of the grass-based livestock husbandry (GLiH; Wezel et al., 2020). An et al. (2021) pointed out that forages played a key role in catalyzing transformation to livestock production, and close nutrient loops. Pereira et al. (2022) measured and monitored the conversion and loss of nitrogen in grazing pastures, and proposed a more accurate nitrogen fertilizer scheme that could be developed in pasture livestock systems supplied by tropical perennial grasses.

From the improvement of the basic elements of grass-livestock interaction in the first period (2000–2011), to the improvement of production capacity and ecological protection in the second period (2012–2016), to the deepening and refining of management strategies in the third stage (2017–2022), research priorities of grass-livestock interaction are discrepant in different periods and different countries, therefore it is necessary to stand in a global perspective, integrate resources and superior scientific research forces to work together so as to solve the global common problems in grass-livestock interaction. New ideas and methods are generated in multiple disciplines (Stock and Burton, 2011), and the joint thinking and cooperation of experts in different fields is conducive to the solution of complex problems concerning grass-livestock interaction (Zhang et al., 2023).

In addition, the global ecological environment is complex and diverse, and the key research points on grass-livestock interaction in different countries and regions may change due to different geographical conditions and needs. For example, Brazil has the most biodiverse tropical savanna in the world, and clarifying invasive alien species and protecting the original grassland structure are the key directions on grass-livestock interaction research (Hilario et al., 2017; Guerra et al., 2020) As one of the most fire-prone ecosystems in the world, the destructive effects of fire management on grassland ecosystems in the Australian inland has received extensive attention (Beringer et al., 2015; Andersen, 2021; Santos et al., 2022). The grass-livestock interaction research of a single country lacks a global perspective and cannot form a joint force since different countries/regions emphasize different research directions. In the future, the research on the grass-livestock interaction should strengthen the cooperation and exchange between different regions and different subject areas, maximize the integration and utilization of resources, so that the research results can be disseminated in a wider range and get more comprehensive and objective practical feedback.

In the face of higher precision data analysis and multi-type and multi-scale monitoring needs (Bhattacharjee et al., 2018; Daly and Fenelon, 2018; Kupková et al., 2023), more detailed and analytically robust indicators and evaluation systems are needed for the grass-livestock interaction study. Remote sensing plays an irreplaceable role as an efficient and efficient analysis tool (O'Connell et al., 2016; César de Sá et al., 2022). According to previous studies, the existing evaluation system of grass-livestock interaction mainly focuses on grassland production, livestock composition, management mode, soil environment and economic benefits, supplemented by drought prediction model and environmental factor analysis (Rao et al., 2015; Pereira et al., 2022; Subhashree et al., 2023). In recent years, the development of multispectral, hyperspectral, infrared, radar, cubesats and other sensing technologies has promoted the use of remote sensing technology in the field of grass and livestock interaction (Zhang T, et al., 2022), and achieved good results (Weiss et al., 2020; Tzanidakis et al., 2023). Future research should develop machine learning and deep learning algorithms (Jiang et al., 2022; Munyati and Mashego, 2023), accompanying with UAV remote sensing (UAV-RS; Diaz-Gonzalez et al., 2022), realize automatic processing and classification based on large-scale data sets (Elkind et al., 2019; Pfitzner et al., 2021; Sun Y, et al., 2022), and establish a remote sensing monitoring system integrating space, earth and space to improve the efficiency and accuracy of data analysis.

Remote sensing technology can cover a large area and obtain more comprehensive image information (Shafique et al., 2022; Wang et al., 2022), including the monitoring of important indicators affecting grass and livestock interaction such as vegetation cover (Hijmans et al., 2005), water content (Hansen et al., 2013) and soil quality (Hersbach et al., 2020), which can capture changes in a short period of time, conduct temporal monitoring of different times and seasons, and track their change trends (Lu and Weng, 2007). It provides an efficient tool for monitoring and managing grazing (Wolf et al., 2021), studying grassland degradation (Tasumi et al., 2014; Hunter et al., 2020), estimating biomass (Ali et al., 2014), and estimating climate change (Cao et al., 2022). In the research of grass-livestock interaction, Dusseux et al. (2015) introduced PaturMata model to evaluate consequences of current environmental changes on grasslands. Zhang et al. (Zhang et al.) established grass production random forest estimation model using the CACART decision tree to evaluate total yield result and the grassland load pressure. However, the existing remote sensing observation technology is still not perfect, mainly spectral images, overlap and occlusion between leaves, different lighting conditions and different growth stages, easy to cause inaccurate monitoring data, coupled with environmental, atmospheric conditions, phenological period and other factors caused by spectral variability, so that the application scenario is limited (Huang et al., 2023). The selection indexes of remote sensing monitoring for grassland, such as coverage, height, biomass and soil organic matter, are not uniform (Zhu et al., 2021) due to different monitoring indicators, reference basis, and selection of research areas and there is a lack of systematic combing of remote sensing monitoring analysis indicators and analysis methods (Sun X, et al., 2022; Zhu et al., 2022).

Grassland sustainability is of great significance in addressing climate crisis (Lei et al., 2016; Abhilash, 2021), enhancing food security (Meli et al., 2023), protecting biodiversity (Fick and Evett, 2018; Yang et al., 2021), maintaining water cycle and carbon sequestration (Overbeck et al., 2015; Török et al., 2021). In response to climate change, with the rise of temperature and the change of precipitation pattern, grassland may face the threat of drought and water shortage (Öztürk and Sen, 2022). Studies such as Wang M, et al. (2023) show that nitrogen deposition, warming and precipitation can change regional vegetation composition, improve grassland productivity, and increase grassland carbon sequestration capacity. To evaluate the productivity of livestock production, a comprehensive consideration should be given to market demand, feeding methods, processing technology and other factors. Lopez et al. (2024) believed that the human feeding system was helpful to avoid overgrazing and promote the regeneration of natural plants, and the adoption of grass feeding can help improve the quality of mutton. Human activities have a huge impact on grassland ecology (Chu et al., 2022; Evans et al., 2023). For example, animal husbandry is the most important economic activity in tropical Mexico, however, grassland has been seriously degraded, and at least 24 states have exceeded grassland carrying capacity due to poor management and continuous expansion of factory agriculture (Quiroz et al., 2021). At the policy and management level, the optimization of management strategies has a positive effect on restoring grassland degradation (Hu et al., 2019; Laitinen et al., 2022; Tao et al., 2022; Dang et al., 2023). Hou et al. (2023) discussed the impact of land tenure reform on grassland quality in China’s pastoral areas based on remote sensing survey data of nearly 40 years, and found that grassland quality increased by about 3% after the privatization of land use rights, but it was still necessary to provide policy support related to environmental safety and environmental protection to expand its positive impact. Furthermore, many countries in the world still need to establish reasonable land management policies and land ownership systems (Achieng et al., 2023), and the allocation of grassland use rights is crucial to protect grassland ecology and maintain grassland sustainability (Barcus, 2018), which is a key issue to ensure the long-term health of grassland ecosystem (Liu et al., 2020; Mangialardo and Micelli, 2021).

Grazing is the most important usage of grassland, and it is the crucial issue to future research on ecological grassland husbandry to clarify the interaction process of grass and livestock under grazing interference and the cooperative regulation mechanism of above and below ground (Tittonell, 2021). How to determine the appropriate grazing density, grazing cycle and grazing area is the main problem in formulating a rational grazing strategy (Hempson et al., 2015). Song et al. (2020) address that free-range grazing is more beneficial to the diversity of intestinal microbiota, and this index can be used as an important marker to improve the evaluation of different grazing management methods. Luo et al. (2023) discover that grassland health assessment is a bridge for grassland ecosystem research, and monitoring and assessment are key components of grazing control (Wang et al., 2022). The health status of grassland can be monitored in real time through remote sensing and geographic information system, which helps to timely adjust grazing strategies in response to different meteorological conditions and ecological changes (de Faccio Carvalho et al., 2021; Machado et al., 2021). A comprehensive review of several studies found that long-term exclusion of grazing could not effectively alleviate grassland degradation, but would lead to a decline in grassland productivity and biodiversity (Zhao and Rokpelnis, 2016). Reasonable grazing system, grazing intensity and timing, and sustainable grazing management can promote the health of grassland ecosystem and maintain grassland ecosystem services (Ding et al., 2020; Kim et al., 2023).

Grass-livestock interaction is a crucial component of grassland ecosystem (Li et al., 2013; Shen et al., 2022), grassland sustainability and grazing control will be the key research directions of grass-livestock interaction in the near future (Figure 8). Ecological, economic, and social factors should be comprehensively considered to ensure the long-term health of grassland ecosystem and the coordinated and sustainable development of animal husbandry (Chu et al., 2022). Future research should integrate advanced science and technology, especially remote sensing technology and machine learning, to improve the efficiency and accuracy of data acquisition (Lutta et al., 2020; Casenave et al., 2022; Hu et al., 2022). Global information on grassland and rangeland resources, establish a more comprehensive indicator system (Toscan et al., 2017; Qiu et al., 2020; Gray et al., 2022), should be conformity and cooperate with the optimization of management policies to provide effective ecosystem assessment and meet the socio-economic needs of sustainable development.