Introduction
Bibliometric analysis is a quantitative observation tool to recognize research gaps in topics of scientific interest such as climate change (Wang, Zhao, & Wang, 2018), development and sustainable livelihoods (Caiado, Dias, Mattos, Quelhas, & Filho, 2017 Zhang, Fang, Chen, & Congshan, 2019), deforestation (Aleixandre-Benavent, Aleixandre-Tudó, Castelló-Cogollos, & Aleixandre, 2018) or analysis methodologies (Chen, Liu, Luo, Webber, & Chen, 2016). The bibliometric analysis allows us to explain the current situation of research topics, development trends and convergence points, providing guidance for future research. In this sense, the bibliometric analysis of models for temperate forest management (MTFM) in the world will provide guidance for potentially valuable research for Mexico.
At the global level, bibliometric analyses related to MTFM have been scarcely addressed. MTFM has been used primarily for estimating timber production, carbon sequestration and prevention of disturbance effects (Law et al., 2018; Soriano-Luna et al, 2018; Urbano & Keeton, 2017), impacts of forest management intensity (Barefoot, Willson, Hart, Schweitzer, & Dey, 2019), effects of fire and wind (Riggs et al., 2015; Wiedinmyer, & Hurteau, 2010), pests (Camacho & Chong, 2015) and climate (Creutzburg et al., 2016; Lundmark et al., 2014). In contrast, the application of models of ecological responses, nutrient dynamics and carbon sequestration, in the long term, has been poorly evaluated due to the difficulty of making observations in different time series and at large geographical scales (Hume, Chen, & Taylor, 2018; Thiffault et al., 2011). However, the need for mitigation, adaptation and management of forests in the face of global change has generated the implementation of technological tools such as forest models to test complex hypotheses related to the understanding of forest management and decision making for sustainable forestry (Kimmins, 2004).
In the case of Mexico, the comprehensive understanding of forest management in relation to primary productivity dependent on the availability of nutrients, organic matter and effects of climate change, is still poorly addressed; in addition, soil and vegetation analyses are made on the basis of single measurements or a single rotation cycle. Consequently, it has not been possible to observe progressive changes and future impacts of management on nutrient availability and biomass production; this is insufficient to establish general conclusions on the effects of forest harvesting in the long term. On the other hand, the use of forest models would facilitate the comprehensive understanding of temperate forests, allowing the simulation of impact or management scenarios, in order to increase wood production without compromising the sustainability of forest (Wang et al., 2014; Zhang, Meng, Bhatti, Trofymow, & Arp, 2008).
Since the MTFM panorama in Mexico is not yet fully understood, it is important to know the emerging trends in research on this topic at a global level, in order to direct the development of relevant research for the improvement of forest management in Mexico. The objective of this review was to recognize scientific advances on the global modeling of MTFM in relation to productivity, carbon storage and nutrient cycling, from the bibliometric analysis of the main thematic areas, emerging research trends and critical points, as well as to focus research of MTFM in Mexico.
Materials and methods
Data collection and processing
The bibliometric analysis was performed by collecting data from the Web of Science Core Collection platform, a documentary database where all contributions (articles, editorials, letters, reviews and discussions) published in science and technology journals indexed by Thomson Reuters are collected. The access to full texts of publications was made by means of a search based on keywords. Keywords were identified according to the most important components of the study objective, in the English language. Search equations were created by mixing keywords (Table 1) and variants (with “*”) with operators (AND, OR, NOT) and reserved symbols (quotes, wildcard character [*,?] and parenthesis). Search included publications from the period 2005-2019. This consultation generated records of titles, keywords, abstracts, institutions, authors and references cited in *.txt format, which were exported to CiteSpace and VOSviewer, for bibliometric analysis.
Bibliometric analysis
CiteSpace (Chen, 2006) and VOSviewer (Wong, 2018) are scientific visualization software packages to identify major thematic areas, emerging research trends and critical point problems. Records previously acquired in *.txt format were linked to VOSviewer and CiteSpace to be analyzed according to "country", "institution" and "keyword" nodes. The identification of the countries providing a topic can help visualize the main factors contributing to the evolution of the knowledge map. Current research topics and frontiers research in the field are identified by the frequency of keywords used in journal articles and the form of collaboration between institutions. Based on the bibliometric analysis, MTFM approaches were considered with the aim of complementing the analysis of research gaps and directing future research. As a result, we obtained recording tables with frequencies and centrality, representative images of nodes and connection networks, where a node represents an element and a pivot point with a high interrelationship centrality.
Results and discussion
Scientific productivity of countries
In the bibliometric review, 960 records were obtained from JCR indexed journals in the MTFM field during the period 2005-2019. The United States was the country with the highest scientific production with 257 articles published, followed by Canada with 102, Germany with 92, China with 88 and France with 60 articles; these five countries accounted for 62 percent of the scientific publications, hosting a large number of authors and academic institutions (Figure 1). Publications and collaborations have increased since 2012, indicating that MTFM is becoming increasingly relevant worldwide, due to the need for mitigation and adaptation of forests to global change. Adaptation of forests to climate change requires innovative methods to integrate a wide range of biophysical and social elements of the system and scenarios of important variables for any management decision (Rastetter, 2017). In this context, the United States, Germany, France, and Sweden have world-renowned research institutes in agricultural sciences characterized by development and application of various models with different objectives and concepts. Modeling has been directed at integrating ecological, economic, and social functions of forests to achieve a multipurpose objective (Pretzsch, Grote, Reineking, Rötzer, & Seifert, 2008). In contrast, Mexico is below the average for global publications; however, from 2015, Mexican publications in indexed journals tripled. This is a result of the role played by forest policies in the country, which have prioritized forest management as a development objective. The forest sector contributes approximately 0.19 % of the national gross domestic product, creating around 177 622 permanent jobs (period 2009-2012), generating economic income and social welfare (Bray, Merino-Pérez, & Barry, 2007; Comisión Nacional Forestal [CONAFOR], 2020).
Information and communication technologies have facilitated the dissolution of institutional, spatial and disciplinary boundaries; in this way, forest researchers, experts or technicians have been able to link up and work with peers at different latitudes. In this sense, the production of grey literature, limited mainly to institutional production, continues to give way to production in scientific journals with greater potential for dissemination and outreach (Aguado-López et al., 2009).
Global network for scientific-institutional collaboration
About 988 institutions have contributed to MTFM research. Among the most active we found The Chinese Academy of Science (CAS), which produced 49 articles with a total of 485 citations; followed by US Forest Service (USFS) with 38 articles and 926 citations; Natural Resources of Canada (NRCan) with 22 articles and 1 076 citations; Institut National de la Recherche Agronomique (INRA) with 20 publications and 1 412 citations; and University of British Columbia (UBC) with 15 articles and 602 citations. This indicates the development and specialization of these research institutes and academies in the field, since 10 % of the GDP of these regions comes from forestry and agricultural production; several of them are governmental organizations. However, the greater number of publications does not necessarily mean the greater influence in the academic field; for example, NRCan, INRA and USFS report greater academic authority in forest management studies, because their articles are cited by a greater number of researchers in the world. Academic production focuses on understanding management practices to sustain forest resources, understanding the effect of disturbances such as climate change, the relationship of forests and human systems, and approaches to forest bioeconomics, bioenergy and bioproducts. In addition, there are institutions that promote innovation and technological advances based on the development of models for management and decision-making in forestry and agriculture. In this sense, the greater collaboration between institutions at the global level is reflected by the greater centrality of collaborative networks (Figure 2): USFS, UBC and CAS were the most influential during the period under review, while Tech Univ Munich showed the most collaboration in 2019, and Swedish University of Agriculture and Science and INRA had the most influence during 2005 and 2007, respectively. The most influenced institutions (centrality) have a higher frequency of publications and citations.
Mexico's scientific production, disseminated internationally, is led by the Universidad Nacional Autónoma de México and the Colegio de Posgraduados. The low centrality at a global level shows the limited collaboration with other institutions in the world, most of which is at a national level; internationally, there is interaction with the University of Wageningen (The Netherlands) and USFS. However, national research institutes stand out in the country; for example, the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP), which leads forestry research and has its own journal "Revista Mexicana de Ciencias Forestales", which publishes much of the work and research on the topic. Therefore, this literature, as well as reports, theses and "grey literature", continues to be relevant for the country, due to the scarce scientific production published in internationally indexed journals. In this regard, there are opportunities and some national strength that could be exploited, such as inter-institutional collaborations that would increase if academic, institutional and international exchange programs were promoted to broaden discussion and facilitate the learning of technologies in the forestry sector, improving indicators of centrality. At the same time, foreign investment and joint projects would be accessed to research Mexico's temperate forests, enabling technology and innovation-based decision-making to be guided. Finally, it is necessary to strengthen the research system in Mexico and the publication in journals that meet quality standards, visibility and impact at the international level. This would be an effective way to increase dissemination and international circulation of the country's research, reflecting science, institutions, researchers and society (Ramírez, Martínez, & Castellanos, 2012).
Research topics on models for temperate forest management
The analysis of the frequency of keywords that summarize the article's topic, areas or methods, allows us to explore research trends effectively and identify “convergence points” in a given topic (Wang et al., 2018). The most frequently keyword in relation to MTFM (Table 2) was “management”, because forest management is based on different forest practices and productive management according to the species and areas where they are employed; furthermore, because management is evaluated or managed to improve aspects of timber production. The word “climate change” was the most common in the global ranking, because it is the main disturbance addressed in modeling due to effects on wood production, biodiversity and incidence on other disturbances (pests, fire and wind), and the need to generate mitigation strategies (sustainable forest management). The words “nitrogen”, “ecosystem” and “carbon” can be considered as main topics guiding modeling research in managed temperate forests based on soil-plant-atmosphere relationship; highlighting the availability of nitrogen for production, storage and sequestration of carbon as the main focus of research. The word “growth” is directly related to wood production, the main objective of forestry production, and “dynamics” is the topic that is understood by models to maintain wood production. Finally, the word “modeling” was used in most of the publications from 2010; it is transversal to all the previous topics, because the aim of modeling is to simulate the influence of disturbances on dynamics of processes and functions to increase wood production or carbon sequestration in managed temperate forests.
Modeling application topics for managed temperate forests
Managed temperate forest modeling has focused on understanding the role of these forests in carbon sequestration and increasing wood production in relation to disturbances (Sharma, Bohn, Jose, & Cropper, 2014; Wang, Bauerle, & Reynolds, 2008). Research applying modeling to estimate or improve wood production uses tree growth models focused on allometric relationships (Bryars et al., 2013; Soriano-Luna et al., 2018; Urbano & Keeton, 2017; Zhang et al., 2008) and relationships with disturbances that negatively impact timber production and cause significant global economic losses (Locatelli et al., 2016); some of them are pests (Camacho & Chong, 2015), fire and wind effect (Riggs et al., 2015; Wiedinmyer & Hurteau, 2010) and climate change (Creutzburg et al., 2016; Dangal, Felzer, & Hurteau, 2014; Harper et al., 2016; Klesse et al., 2018; Lundmark et al., 2014; Wang et al., 2014). In the context of climate change, models have focused on the design and implementation of mitigation and adaptation policies (Wang et al., 2013).
Scientific production in forest modeling focuses primarily on the relationship between carbon sequestration and storage capacity (Ricker, Gutierrez-Garcia, & Daly, 2007; Thom, Rammer, Garstenauer, & Seidl, 2018; Woodbury, Smith, & Heath, 2007). For example, C emissions from forest conversion to managed stands, and type of disturbance and intensity determine C losses to the atmosphere (Chen et al., 2013). Disturbance and nutrient cycling are also analyzed by flows models and interactions between forest processes, in order to develop more realistic predictions of forest response to management practices and global change (Karam, Weisberg, Scheller, Johnson, & Miller, 2013). In addition, estimates of spatial and temporal changes in the loss of temperate forest cover allow us to estimate vulnerability to global climate change (Potapov, Hansen, Stehman, Loveland, & Pittman, 2008). The role that these forests play in the carbon cycle at stand and landscape scale has tried to be understood (Gonzalez-Benecke, Martin, Cropper, & Bracho, 2010; Law et al., 2018; Manzoni & Porporato, 2009; Thompson et al., 2016), complementing with assessments on carbon management strategies based on forest harvesting life cycles (Peckham & Gower, 2013; Winford & Gaither, 2012). Modeling has been used in smaller-scale studies to understand how limiting resources influence forest production, such as water deficit and nutrient availability (Griffiths et al., 2019; Liu et al., 2018; Seidl, Rammer, Jäger, Currie, & Lexer, 2007; Tian, Youssef, Skaggs, Amatya, & Chescheir, 2012), and processes related to harvesting and soil carbon and nitrogen, such as organic matter stabilization, decomposition processes and microbial response to root exudates (Abdelnour, McKane, Stieglitz, Pan, & Cheng, 2013; Bhowmik et al., 2017; Parolari & Porporato, 2016; Robertson et al., 2018; Smethurst et al., 2015; Wallace, Laughlin, Clarkson, & Schipper, 2018; Wang et al., 2014).
Modeling provides biological and analytical realism for understanding the structure and function of forests; however, the complexity of some models adds uncertainty to the underlying causes of final predictions, weakening their heuristic value or providing useful theoretical information but omitting critical details (Kimmins, 2004). The bibliometric analysis identified the main research gaps that need to be addressed to improve ecosystem management; for example, how the use of soil-plant-disturbance relationship models can integrate independent biodiversity factors, drivers of change and ecosystem responses to help improve assessments of managed forests and their sustainability.
Modeling Research Guidance for Managed Temperate Forests in Mexico
In Mexico, most of the research that has used MTFM is related to aerial biomass allometric estimates (Chávez-Pascual, Rodríguez-Ortiz, Enríquez-Del Valle, Velasco-Velasco, & Gómez-Cárdenas, 2017; Corral-Rivas et al., 2017; Douterlungne, Herrera-Gorocica, Ferguson, Siddique, & Soto-Pinto, 2013), carbon expansion equations of several plant species (Silva-Arredondo & Návar-Cháidez, 2009) and spatial equations to determine impacts of land cover change in relation to management and carbon sequestration (Flamenco-Sandoval, Ramos, & Masera, 2007; Prieto-Amparán et al., 2019; Ricker et al., 2007; Soriano-Luna et al., 2018). However, there are gaps in the comprehensive understanding of forest management in relation to primary productivity dependent on availability of nutrients, organic matter and climatic factors, and forest management.
Currently, forest management impacts positively on ecosystem services of timber production (raw material), carbon sequestration and storage, and affects in a negative way the services of conservation of plant diversity and regulation of water flows (Monárrez-González, Pérez-Verdín, López-González, Márquez-Linares, & González, 2018). In this sense, the transition from a purely silvicultural vision to a systemic vision would allow the management of forests from the understanding of the compensations, synergies and relations of ecosystem services (Galicia & Zarco-Arista, 2014). Research based on modeling would allow the integration of elements and simulate diverse disturbance scenarios to generate forest management that is indeed on the way to sustainability. For example, it is necessary to model the impacts of biomass harvesting on vegetation responses (richness, composition and diversity of plant species); furthermore, soil disturbance generates other less visible consequences in ecosystems such as the modification of microbial communities and, consequently, of soil functions, such as the interdependence of C and N cycles (Nasi & Frost, 2009) which have been little explored. On the other hand, modeling would allow identification, quantification and valuation of multiple ecosystem goods and services in sites under forest exploitation, for the application of social, economic and environmental policies on long-term ecosystem management strategies (Galicia et al., 2016); for example, the multifunctionality of mixed plantations depends greatly on the arrangement and appropriate combination of species for the achievement of objectives. In this sense, emphasis has been placed on stimulating productivity and growth of forests through mixtures of species (evergreen-broadleaf) and it has been determined that these mixtures provide the greatest ecological benefits (i. e. maintenance of fertility) for conservation, protection and restoration, in comparison with monospecific pine forests, although these have a higher commercial value (Nunes, Lopes, Castro, & Gower, 2013). Therefore, simulations of ecological processes allow proposals to be made for mixed plantations to increase biomass production, C storage in the soil and availability of nutrients in relation to monospecific plantations (Forrester, 2014). In addition, modeling is a key to the successful implementation of mitigation practices, as it requires knowledge of the role of species identity and diversity in the long-term accumulation of C in plantations.
The use of models in the context of forestry in Mexico could help us understand the consequences of management in the Mexican method of irregular forest management and the method of forestry development, integrating aspects such as regeneration and maintenance of soil characteristics and biodiversity. In other words, the use of models would answer the following questions: How do forest management systems affect structure and composition of plant communities? How has the type of forest management affected nutrient stores and dynamics? What are the temporal dynamics of emissions of major greenhouse gases under different management? and How does forest management affect the ability to maintain biodiversity, carbon sequestration and forest soil fertility? The use of models opens the possibility of solving the above-mentioned problems, given that in our country the scientific information that supports forest management decisions, both operationally and normatively, is limited.
The application of MTFM in the country would require improved data collection through standardized protocols for sampling, sample analysis and management of these ecosystems as a starting point for simulating disturbance effects in conditions and contexts of the regions of Mexico. Moreover, since the number of forest cycles employed in the country varies between 10 and 50 years, it is important that research focus on developing design criteria to improve the configuration of tree species and density of plantations and benefits, in the long term. In this context, agroforestry statistics should be collected and updated at the institutional level, so that existing models can be calibrated and validated by various researchers in the country; for example, plant-soil relationship and forest management intensity models could be used such as EFIMOD (Mikhailov et al., 2003), which models carbon and nitrogen; CENTURY simulates, in addition to the above mentioned elements, phosphorus and sulfur (Parton, McKeown, Kirchner, & Ojima, 1992); Forest DNDC (Li, 2000), Biome-BGC and Forest-BGC (Aber & Driscoll, 1997) simulate carbon, nitrogen and water; CBM-CFS3 (Kull et al., 2011) models carbon emissions and life cycle of wood products; PnET-BGC, PnET-CN (Svensson, Jansson, & Kleja, 2008) simulates energy, water and major element flows; and COUP MODEL is a coupled heat and mass transfer model of the forest ecosystem. All of these models allow for testing management or climate change scenarios in relation to their main analysis. In this way, MBTM makes it possible to identify the effects of disturbances or management on carbon sequestration, nutrient availability, primary productivity and water availability, which in turn could reduce disturbances in wood production due to natural and human-induced hazards.
Conclusions
The bibliometric analysis suggests that the United States is the leader in research on models for managed temperate forests (MTFM). The US Forest Service, Natural Resources of Canada (NRCan) and the Institut National de la Recherche Agronomique (INRA) report the greatest academic authority, because their articles are the most cited in the world. Publications and collaborations have increased since 2012 indicating that MTFM is becoming more relevant, because of the need for mitigation and adaptation of forests to global change. Mexico and other Latin American countries have less centrality in publications and collaboration at a global level in this topic because the Web platform of Science Core Collection only collects contributions from indexed science and technology journals. Therefore, it is important to promote interpersonal and inter-institutional collaboration to improve the appropriation of technologies, foreign investment and increase the number of indexed publications. Finally, it is necessary for MTFM in Mexico to move towards an integral analysis for sustainable management.