氮磷钾施肥对农业的环境效应:文献计量分析与综述(原标题:Environmental Effects of Nitrogen, Phosphorus, and Potassium Fertilization in Agriculture: A Bibliometric Analysis and Literature Review)

1. Introduction

Agriculture underpins global food security; yet, the widespread and inefficient use of nitrogen (N), phosphorus (P), and potassium (K) fertilizers has placed substantial pressure on environmental systems [1,2,3]. These pressures manifest as water eutrophication, soil degradation, greenhouse gas emissions, air quality impacts, and effects on biodiversity and ecosystems, which collectively constitute the “agricultural environmental effects” examined in this work. The agricultural environmental effects of these nutrients are frequently interrelated rather than isolated [4,5,6]. For example, the synergistic N-P interactions can intensify water eutrophication, while K deficiency can limit the crop uptake of N and P. As essential nutrients, N, P, and K both enhance crop production, and, when mismanaged, generate a cascade of environmental effects that require systematic analysis [7,8,9,10].

The existing research on the effects of nitrogen, phosphorus, and potassium fertilizers on the agricultural environment can be summarized into the following three aspects. First, the irrational use of nitrogen leads to a range of environmental challenges worldwide [11]. Especially in the production of major food crops such as rice, wheat, and corn, excessive nitrogen fertilizer is the main source of ammonia (NH 3) emissions, posing a threat to air quality and human health [12,13,14]. Taking China as an example, the excessive application of nitrogen fertilizer has caused serious problems such as the eutrophication of water bodies, soil acidification, and increased greenhouse gas emissions [14,15]. Studies have shown that optimizing fertilization management can significantly reduce nitrogen loss and improve utilization efficiency, thus alleviating its negative impact on the environment [15,16]. Secondly, phosphorus is essential to agricultural production, but the agricultural environmental effects caused by its improper management cannot be ignored [5,17,18,19]. For example, in intensive breeding areas, if the high content of P in livestock and poultry manure is not properly treated, it is easy to pollute the surrounding farmland and affect agricultural systems and food security [18]. In addition, K is a key nutrient element for plant growth, but its excessive or improperly managed application raises distinct environmental concerns that differ fundamentally from those of N and P [20,21,22]. It should be noted that, compared with the extensive literature on the environmental effects of N and P, the direct environmental impacts of K fertilization remain considerably less studied, and most of the existing research addresses K in the context of its synergistic or antagonistic interactions with N and P rather than as an independent environmental stressor [6].

Although the individual environmental effects of NPK are well-documented, the current knowledge base has critical gaps that call for an integrated assessment. Most studies focus on single nutrients or isolated environmental endpoints, yielding fragmented insights that overlook the interconnected impacts of N, P, and K interactions [4,5,6]. While the general awareness of the NPK-related environmental harm is widespread, no systematic, quantitative mapping, and thematic evolution has yet been conducted specifically for agricultural NPK effects. For instance, P and K shortages can restrict N use efficiency, thereby reducing the fertilizer effectiveness and exacerbating environmental pollution [5,6]. Significant gaps also persist in understanding integrated NPK interactions under changing climatic conditions, compound risks from multiple stressors, and the links between scientific evidence and governance systems [7]. Thus, a comprehensive evaluation of agricultural environmental effects must account for the synergistic mechanism of N, P, and K. Moreover, no previous study has proposed a unified conceptual framework that integrates fertilizer inputs, NPK interactions, soil processes, environmental outcomes, management practices, and policy support—a gap that this review precisely aims to fill.

A considerable body of review literature already exists on specific environmental processes and individual nutrients related to agricultural fertilization. Process-specific reviews have examined P-driven eutrophication and N–P synergistic effects on water quality [4,17], N–P non-point source pollution control, greenhouse gas emissions, and mitigation strategies in agricultural systems [15,22,23]. Nutrient-specific reviews have synthesized the environmental fate of nitrogen, P solubilization and loss management, and K availability and crop responses [18,20,22]. Additionally, several bibliometric studies have mapped the research trends related to agricultural greenhouse gas emission reduction, carbon neutrality, and agricultural management [23,24,25,26,27]. However, these existing reviews share a common limitation: they address either a single environmental process (e.g., eutrophication or GHG emissions) or a single nutrient (N, P, or K) in isolation. No prior study has combined all three nutrients across multiple environmental processes within a unified analytical framework—synthesizing the bibliometric mapping of the research landscape with a literature review that identifies cross-cutting knowledge gaps and proposes a comprehensive management framework. This gap is significant because the environmental effects of N, P, and K are fundamentally interconnected through synergistic and antagonistic mechanisms, and understanding these interactions demands a holistic approach that transcends the boundaries of process-specific and nutrient-specific reviews. Addressing this gap will advance our understanding of the evolution, knowledge structure, emerging research priorities, and future directions of the agricultural environmental effects research related to N, P, and K.

Accordingly, this study uses a combined bibliometric and literature-review approach with four specific objectives: (1) to quantify the temporal publication trends and the geographical, institutional, journal, and disciplinary structure of the field; (2) to identify the dominant and emerging research themes through citation, co-citation, keyword co-occurrence, clustering, and related bibliometric analyses; (3) to synthesize substantive evidence on N, P, and K cycling, environmental effects, management technologies, and policy responses, explicitly using the bibliometric results to guide the literature discussion; and (4) to identify cross-cutting knowledge gaps and develop evidence-informed priorities and an integrated framework for sustainable nutrient management. The bibliometric analysis provides the quantitative mapping presented in Section 3, whereas Section 4 interprets and synthesizes the substantive literature in relation to those mapped patterns.

The overall objective is therefore to map the evolution, knowledge structure, and emerging frontiers of research on the environmental effects of NPK fertilization in agriculture, and to translate those patterns into clearly defined knowledge gaps and strategic priorities for sustainable nutrient management. The remaining sections of the paper are organized as follows: Section 2 provides the methods and data source; Section 3 provides the bibliometric analysis; Section 4 presents the literature review; and 5 gives the conclusions, respectively.

2. Data and Methods

Literature data were collected on 10 March 2025 from the Web of Science (WoS) Core Collection, specifically the Science Citation Index Expanded (SCIE) and the Social Sciences Citation Index (SSCI). The combined use of SCIE and SSCI was intended to capture both natural-science and social-science dimensions of agricultural environmental-effects research. The retrieval formula was as follows: TI = (agriculture* OR Agro* OR husbandry) AND TS = (Nitrogen* OR Phosphorus OR Phosphor OR potassium OR Kalium OR NPK). Here, TI (Title) restricts the agricultural-context terms to article titles, whereas TS (Topic) searches titles, abstracts, author keywords, and Keywords Plus, thereby providing broader coverage for nutrient terms. The wildcard “Agro*” retrieves terms such as agronomy, agroecosystem, agroecology, agroforestry, and agrochemistry. Because this broad term can also retrieve studies outside the intended scope, all records were subsequently screened manually. The initial search yielded 3756 records from 1991 to 2024. Inclusion criteria were as follows: (1) indexing in SCIE or SSCI; (2) document type of article or review; (3) English language; (4) publication year 1991–2024; and (5) substantive relevance to environmental effects associated with N, P, or K fertilization in agricultural systems. Exclusion criteria were as follows: (1) records unrelated to agricultural environmental effects, including studies centered on brain research, drug development, DNA sequencing, medicine, or other off-topic areas; (2) non-English publications; (3) document types other than articles or reviews; and (4) duplicate records. Relevance screening and exclusion of off-topic records were conducted manually by reviewing titles and abstracts rather than by automated keyword deletion. To provide an external check on search recall, the 20 most-cited records retrieved by the search were cross-referenced against the reference lists of three relevant reviews; all 20 were captured by the search strategy.

Data cleaning was then performed to improve consistency in country and institutional attribution [28,29]. Country-name variants representing the same analytical national unit were consolidated before network and publication analyses. Records from mainland China, Taiwan, Hong Kong, and Macao were grouped under China, while records from England, Scotland, Wales, and Northern Ireland were grouped under the UK. This harmonization was applied consistently to avoid splitting publication and network counts across multiple geographical labels used in WoS records; the resulting country-level indicators should therefore be interpreted according to these analytical groupings. Institution-name variants were similarly standardized using DDA. After application of the inclusion/exclusion criteria and cleaning procedures, 3649 papers were retained for analysis. In addition, 44 records were identified as “Highly Cited Papers” using the WoS designation based on Clarivate’s Essential Science Indicators (ESIs). Under this system, a Highly Cited Paper ranks within the top 1% by citation count for its ESI field and publication year, providing a field- and time-normalized indicator of citation performance.

Two software tools with complementary analytical functions were employed. VOSviewer version 1.6.16 (Leiden University, Leiden, The Netherlands; https://www.vosviewer.com/, accessed on 25 March 2025) was used for the following: (a) co-authorship analysis (countries and institutions); (b) keyword co-occurrence analysis and clustering visualization; (c) journal co-citation analysis and mapping; and (d) bibliographic coupling analysis. Derwent Data Analyzer V7 (DDA, Clarivate Analytics, Boston, MA, USA) was used for the following: (a) data cleaning and consolidation (merging country and institution name variants); (b) publication trend analysis; (c) country and institution distribution analysis; and (d) subject category distribution analysis. The combined use of these three tools ensures both analytical depth and visual clarity while reducing tool-specific bias. The research workflow of this study is illustrated in Figure 1.

Figure 1. The research workflow of this study.

3. Bibliometric Analysis

3.1. Publication Trend

From 1991 to 2024, the research of agricultural environmental effects developed rapidly, which was fully reflected from the number of articles. An analysis of the publication trend shows that the cumulative number of articles was 3469, with annual publications ranging from 15 to 290 during 1991–2024 (Figure 2). Annual publications ranged from 15 to 290, and the fitted exponential curve (R 2 = 0.9661) indicates an overall accelerating publication trend, despite the year-to-year fluctuations. The first stage was during 1991–2005, with a cumulative 569 publications (the annual number was 38), being 16.40% of the total. The second stage was during 2006–2024, with a cumulative 2900 publications (the annual number was 153), accounting for 83.60% of the total. From the perspective of the overall trend of the published articles in the agricultural environmental effects research, although there has been a downward trend in some years (such as 1996, 2003, et al.) comparing to the previous year, it showed varying degrees of growth trends over last 34 years. The results revealed that the research field of agricultural environmental effects has been regarded widely with concern by academic circles and society in recent years.

Figure 2. Number of articles in the agricultural environmental effects research from 1991 to 2024.

3.2. Analysis of Country Distribution

Figure 3 depicts the number of articles published in agricultural environmental effects research by the top 10 countries, including the USA, China, and Canada, during two distinct time periods, 1991 to 2005 and 2006 to 2024, as well as their proportions of the total publications. From the changes observed in the number of published articles, China experienced a notable increase in the number of articles from 2006 to 2024, surpassing the level recorded from 1991 to 2005. This trend showed China’s rapid development in this research field. This sustained expansion is quantitatively evidenced by the pronounced increase in its average annual publication volume: the annual output of peer-reviewed papers climbed from less than 50 in the 1991–2005 period to more than 150 throughout 2006–2024. In contrast, although the USA maintained a substantial number of articles in both periods, its growth rate remained relatively modest. In other countries, such as the UK and Germany, the number of articles exhibited relatively steady changes across the two periods. From a proportion standpoint during 1991–2005, the USA held a relatively high proportion of research published articles in this field, whereas China’s proportion was exceedingly low. By the period spanning 2006 to 2024, China’s share had increased significantly, reflecting its growing influence in international research. Meanwhile, the proportions of other countries either declined to varying degrees or remained relatively stable.

Figure 3. Number of articles in the agricultural environmental effects research from top 10 countries.

Figure 4 presents the national collaboration network on the agricultural environmental effects research from the (a) total publications and (b) total citations (Figure 4). As shown, the nodes represented various countries, while the lines connecting the nodes illustrated the associations between countries. From the network diagram of total publications (Figure 4a), the USA and China occupied central positions in the chart and had more connections with other countries, suggesting extensive collaboration with other nations in this research field. Other countries, such as the UK and Germany, were also connected to numerous countries, highlighting the prevalence of international collaboration in this research field. From the network diagram of total citations (Figure 4b), China and the USA continued to hold relatively core positions, with dense connections to other countries. This suggested that the research outcomes of these two countries were wi