Resilience studies on agroecosystems

In recent years, the term resilience has taken on importance and has been used in various fields of knowledge. In the case of agroecosystems and resilience (as keywords), there have been 188 publications since 1993, with 2018 and 2019 being the years with the most articles published with 30 and 32 articles respectively (Figure 1).

Figure 1 Publications on agroecosystem resilience registered on Web of Science from 1993 to 2020. 

The articles on agroecosystem resilience analyzed based on their frequency of citations were integrated into 11 domains described below: environmental sciences and ecology (71), agriculture (52), environmental sciences (44), ecology (32), science and technology (26), multidisciplinary agriculture (23), green and sustainable science (20), environmental studies (19), biodiversity and conservation (10), soil science (9), agronomy (7), meteorology and atmospheric sciences (4), multidisciplinary science (3) and water resources, plant science, environmental engineering, biology, biomedicine, forestry and engineering with two publications respectively.

According to connectivity based on the intermediate centrality value, the areas that relate to others in greater frequency were mainly environmental sciences and ecology (0.37), agriculture (0.32), environmental studies (0.29), science and technology and other topics (0.27), ecology (0.2) and environmental science (0.14) (Figure 2). The marked relationship between the various areas presents interactions and feedbacks between social, economic, political and ecological problems different space-time scales (^{Castillo and Velázquez, 2015}).

Figure 2 Main disciplines that address resilience in agroecosystems according to the CiteSpace tool. 

Studies have shown that nodes with high intermediation centrality values tend to identify boundary expansion potentials that can lead to transformative discoveries.

The country of origin of the co-authors who have published on resilience are mainly from: United States of America (34), France (20), Australia (10), Canada (7) and Germany (6). With France and the United States of America being the nodes that connect and influence two or more large groups (Figure 3), this is indicated by the size of each node (^{Chaomei, 2014}). The results shown in the networks show similarity in the domains analyzed by ^{Hosseini et al., (2016)}, who, when conducting a literature review on systems resilience, determined the presence of Psychology as the main area and followed by environmental sciences and ecology as areas from which this concept has also been addressed.

Figure 3 Main countries of origin and value of intermediate centrality of co-authors related to the publications of resilience of agroecosystems in the period 1993-2020. 

In relation to the most frequently cited authors of the 188 articles, the next stand out: Altieri (53), Folke (31), Walker (30), Tilman (29), Holling (26), Gliessman (22), Gunderson (16), Kremen (15), Lal (15), Tscharntke (14), FAO (14), Scheffer (14), Lin (12), Carpenter (12), Adger (11), Van Der (11) and Perfecto (11). Regarding the centrality based on the connectivity of the authors through co-quotations, the next stand out mainly: Altieri (0.24), Hooper (0.20), Walker (0.21), Holling (0.12), Folke (0.11), Swift (0.11), Tilman (0.1), Tscharntke (0.08), Lin (0.06) and Koohafkan (0.06). Subsequently, other authors with a lower degree of centrality appeared and these are: Gliessman (0.04), Gunderson (0.04), Scheffer (0.03), FAO (0.03), Lal (0.02), Carpenter (0.02) and Kremen (0.01).

The intensity of the relationship of outstanding authors and those with a lower degree of centrality is due to the number of articles cited that have these authors in their references (Figure 4). In a broader context on resilience of systems, ^{Hosseini et al. (2016)} characterized the presence of authors who study resilience from qualitative, semiquantitative and quantitative approaches.

Figure 4 Network that shows the main authors cited in the 188 articles analyzed. 

Of the total number of articles (188) for the period 1993-2020, and according to Web of Science, the articles most cited in the extensive database and without delimiting their relationship to agroecosystems, these are related to the study of the agroecosystem resilience from the ecological approach (Table 1).

Conceptualizing resilience in the context of agroecosystems

The word resilience comes from the Latin resillire that means to jump backwards and was initially used in the area of physics to describe the stability of materials and their resistance to external impacts (^{Holling, 1973}; ^{Reid and Boterill, 2013}). In the 60’s with the growth of system thinking, the concept was used in ecology (^{Reid and Boterill, 2013}) and its definition highlights two aspects; the first, about the recovery and return to the initial state of the ecosystem after an impact and the second, about how a system has been disturbed and remains without changing its function (^{Bodin and Wilman, 2004}; ^{Masterson et al., 2014}).

Subsequently, the concept was incorporated into agroecosystems and in the late 80’s and 90’s, it was applied within the ecological theoretical framework, particularly taking the foundations of the concept of ^{Holling (1973)}. Returning to the analysis of networks in Figure 4 and based on definitions of the authors with greater centrality, two types of approaches to the study of resilience were identified: an agroecological approach and another approach to socioecological systems.

From agroecology and with the aim of developing sustainable agroecosystems with less environmental impact, highly productive and economically viable, emphasis was placed on resilience as an emerging property of the agroecosystem based on principles such as productivity, stability, sustainability, equity and ecological principles (^{Conway, 1987}; ^{Thomas and Kevan, 1993}; ^{Gallopín, 1995}).

In addition to the above, resilience was conceptualized as an attribute of the health of the agroecosystem and was also studied from the relationship with sustainable practices, environmental services and biodiversity (^{Okey, 1996}; ^{Altieri 1999}; ^{Tscharntke et al., 2005}). The author Walker (2004) integrated into the study of resilience the concept of socioecological systems, which extrapolated to agroecosystems integrates the emerging properties of the producer in their interaction with elements of adjacent systems.

The concept of socioecological systems highlights cyclical adaptability as the ability of the elements of the agroecosystem to influence resilience through constant feedback that leads to building collective structures (based on feedback) that increase their resistance and probability of recovery from disturbances, thus generating more information that is incorporated back into the elements of the system (^{Davoudi, 2012}; ^{Titonell, 2013}).

However, the macro and microsystem are also considered, because in them there may be variables conductive to events of drastic disturbance that could exceed the resistance of the agroecosystem, avoiding the emergence of resilience and instead, a transformation scenario appears (Walker, 2004; ^{Folke, 2006}; ^{Gunderson, 2008}; ^{Folke et al., 2010}; ^{Folke, 2016}).

The extension of notions of resilience in society has important limits, particularly its conceptualization of social change. This is due to the lack of attention to normative and epistemological problems that underlie the notion of social resilience.

So, the role of knowledge at the intersections between environmental and social dynamics helps to address normative issues and their respective values as an integral part of the development and functioning of the socioecological system. Therefore, it is necessary to develop methods of sociology that permit to strengthen the study of human beings in relation to their behavior under stress scenarios caused by external phenomena (^{Cóte, 2011}).

It is likely that unified methods for the study of resilience in agroecosystems are not yet defined. This coincides with a series of definitions on the resilience of agroecosystems and with proposals for variables with which this phenomenon could be addressed.

Components of resilience

The definition of resilience is confusing because of the disciplinary connotations in which it has been used. From a socioecological systems approach, resilience is a component of the agroecosystem trajectory (Figure 5) and is operationalized with four components proposed by Walker et al. (2004): latitude. It is defined as the maximum amount that can be changed to the system before losing its recoverability. Basically, it is the width of the basin of attraction. A wide basin means that the system can experience a greater number of states without reaching the transformation, resistance. It is the ease or difficulty to change the system.

Figure 5 Limits of a socioecological system to express resilience, where: L= means length, R= resistance and Pr= precariousness (Walker et al., 2004).  

It is related to the topology of the basin. Deep basins of attraction indicate that greater forces or disturbances are required to change the current state to another, precariousness. It is the current trajectory of the system, which places it with respect to a limit or ‘threshold’ that, if exceeded, hinders or makes it impossible to recover (Pr), panarchy. Process of influence on the system caused by the states and dynamics of the (sub) systems at scales above and below the scale of interest (Pa), (Walker et al., 2004).

In the community context, ^{Masterson et al. (2014)} propose three dimensions within the resilience process: 1) resistance or robustness is the ability of the capital system to absorb, change or resist the impacts of natural phenomena, which implies a reduction in probabilities of system failures; 2) dimension called speed refers to the recovery path and the speed with which the system recovers its pre-impact state; and 3) the third refers to increasing resilience through learning or adaptability.

^{Twigg (2007)} proposes various dimensions or components for social systems (governance, risk assessment, knowledge and education, risk management and vulnerability reduction, disaster preparedness and response) that a community must assemble to increase its resilience, prioritizing strategies that enable systems to act on their own and strengthen their capacities rather than focusing on their vulnerability to disaster or their needs in an emergency.

From a social perspective, the sustainability and resilience of a community depend on its ability to access and make the best use of its economic, political and civil capital in the face of a disturbance (Akter and Mallick, 2013). Other important factors in the dimensions of resilience are: autonomy, food security, adequate levels of nutrition, health and education to name a few, as well as the versatility of modifying the structure of their system as required in disaster scenarios (De La Fuente, 2016).

Resilience indicators

Any resilience study must undoubtedly start from the questions, what makes an agroecosystem resilient? what are those attributes of the system that give it greater or lesser resilience? and what indicators are intrinsic and extrinsic to the agroecosystem? However, there is likely to be confusion about what aspect of resilience is being measured and what kind of indicators are being considered, since resilience may occur in a later period of time or on a particular scale at the expense of the resilience of the system.

To avoid such confusion, it is important to be clear about the scale and period of time in which the study of resilience is approached (^{Carpenter et al., 2001}; ^{Hernández et al., 2002}). To facilitate the analysis, the outstanding indicators presented in different studies were grouped into seven categories: socioecological systems (^{Cabell and Oelofse, 2012}), agroecological systems (^{Bruneau et al., 2003}; ^{Altieri and Nicholls, 2012}; ^{Altieri and Nicholls, 2015}), social systems (Rose, 2007; Rose, 2009; ^{Cabell and Oelofse, 2012}; CENAPRED, 2015; ^{Hosseini et al., 2016}), community capacities (Rose, 2007; ^{Masterson et al., 2014}; CENAPRED, 2015; ^{Qasim, 2016}), economic capacities (Rose, 2007; Rose, 2009), institutional capacities (Rose, 2007; CENAPRED, 2015; ^{Qasim et al., 2016}) and infrastructure capacities (Rose, 2007; CENAPRED, 2015; ^{Qasim et al., 2016}).

Each category relates to the agroecosystem directly or indirectly and its influence determines the response of the agroecosystem to a crisis. Finally, once the resilience indicators have been selected, it is important to standardize the values obtained for which use can be made of equation (1) proposed by ^{Hahn et al. (2009)}: Iv=Ia-IminImax-Imin 1), where I_v= is the standardized value of the indicator, I_a= is the value for a particular agroecosystem, I_min= for the minimum value of total agroecosystems, I_max= for the maximum value of total agroecosystems (^{Hahn et al., 2009}).

In terms of disaster prevention, it is substantial to focus on increasing the robustness of the agroecosystem and decreasing vulnerability, because these variables are inversely proportional to resilience. An agroecosystem with low vulnerability might probably suffer no or minimal damage.

In the study and design of resilient agroecosystems, it is important to approach allopoietic processes to know how the dynamics of interaction with the environment or external systems are as promoters of changes in the state of the system or internal self-organization.

In the adaptation stage (prior to the expression of resilience), the agroecosystem modifies or changes some elements of its structure in response to the environment around it, so every time this phase ends and an agroecosystem is resilient, it can mean three things: 1) that the agroecosystem presents an internal self-organization; 2) that resilience is an emerging behavior; and 3) that it is a change of state resulting from an agroecosystem with high entropy (^{Dekkers, 2015}). Finally, interdiscipline should be promoted as a methodological tool to address resilience.

This represents a common conceptual and methodological framework, derived from a shared conception of the science-society relationship, which will permit defining the problem to be studied under the same approach, without neglecting the specialization of each of the members of the research team (^{García, 2006}).