A Method to Assess Agroecosystem Resilience to Climate Variability

Abstract

Agroecosystems are influenced by climate variability, which puts their productivity at risk. However, they tend to maintain a functional state through their resilience. The literature presents several methods for assessing general resilience, but for specific resilience to climate variability, there are very few methods. An index is proposed that assesses the resilience of agroecosystems to climate variability, based on approaches and indicators that consider the interrelationships of agricultural systems with the environment. The index is made up of a set of multidimensional indicators, which give weight to the role that these play in the resilience of an agroecosystem. As a result, decision-making is assisted in the attempt to adapt or modify components of a farm, technology, and the culture of farmers. This index conceptually introduces structural and linkage indicators that assess ecological connections within farms and between farms and their environment. To demonstrate the effectiveness of the method, an application was implemented to evaluate the resilience to climate variability of fifty-one farms, located in Colombia, dedicated to citrus production, and it was verified that the most resilient farms were those that have the best qualified indicators, as well as being the ones with the highest level of production and profitability.

1. Introduction

Climate variability and change can affect agroecosystems and thus agricultural production, even in high-yield and high-tech areas [1,2,3,4,5,6], endangering food safety [7,8] and disproportionately affecting the well-being of the poor and those with limited access to land, modern agriculture, inputs, infrastructure, and education [9]. In response, agroecosystems tend to be adapted and transformed to ensure and guarantee system functions [10] through resilience [11], which, in recent decades, has been the target of numerous scientific studies because of the deterioration of the environment [12].

Resilience is an emergent property generated in complex systems, which allows them to cushion, adapt, innovate, and transform in the face of specific stress factors and the inevitable and continuous biophysical and social changes in the environment [13,14,15,16]. It can be interpreted as “the amount of disruption needed to transform a system from one stability domain to another” [17]. “Resilience includes impacts and taking advantage of opportunities” [18] (pp. 300–301). “It is a system’s ability to restore viable evolutionary conditions for a specific course of tychastic uncertainty that is measured in terms of duration ” [19] (p. 251). “In this perspective, resistance to a disturbance and the speed with which a system returns to equilibrium is the measure of resilience. The faster the system bounces back, the more resilient it is” [14] (p. 300).

In ecological studies, “resilience determines the persistence of relationships within a system and is a measure of the ability of these systems to absorb changes of state variables, driving variables, and parameters, and persist is the ability of a system to return to an equilibrium state after a temporary disturbance” [20] (p. 17). Resilience is also applicable to social systems, where it is defined as the ability of groups or communities to interact with external disturbances because of social, political, and biophysical changes [18,21,22].

According to ref. [23], resilience can have two categories: (i) inherent resilience, which depends on the natural characteristics of the system, and (ii) acquired resilience, built on physical, biotic, social, economic, and cultural adaptations, where “human activities have reduced the resilience of managed ecosystems over time, making them vulnerable to disturbances” [17] (p. 247). In this sense, ref. [10,24,25] defined two types of resilience: (i) general resilience, which does not encompass the part of a system that could cross a threshold, nor the types of disturbances or shocks that the system must withstand; (ii) specific resilience: applied to problems that may arise from a specific set of sources or disturbances in certain aspects of a system, for example, climate variability.

In a study on agroecosystems, ref. [24] proposed that resilience is related to three complementary properties: (i) The amount of change that the agroecosystem can undergo and still maintain control of function and structure. (ii) The degree to which the system is capable of self-organization. (iii) The ability to build and increase the capacity for learning and adaptation [10]. This confirmed that resilience is related to the ability to ensure and guarantee system functions in the face of economic, social, environmental, and institutional disturbances through robustness, adaptability, and transformability. It allows agricultural systems to balance the ability to be efficient in each context with the ability to reorganize and adapt in response to unforeseen and unpredictable changes [25].

In the context of climate variability, ref. [26] (p. 1108) defined resilience as the “ability of a social ecological system and its components to anticipate, reduce, accommodate, or recover from the effects of a hazardous event or trend [weather related] in a timely and efficient manner”. In the same context, ref. [27] argued that resilience in agroecosystems is generated when people and communities interact to face the effects of this variability.

From the point of view of the methods and indicators for the evaluation of the resilience of agroecosystems, the literature reports different approaches in local and global territorial contexts [10,11,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44], most of which are related to the analysis of so-called general resilience [10]. This evaluation in developing countries has been approached with indices that seek to capture the greatest number of variables. However, these methods present limitations from a conceptual and methodological point of view, especially with moderate incorporation of social and cultural indicators and the analysis of the interaction of components in an agroecosystem without weights. Additionally, they have not been designed to specifically assess the resilience of agroecosystems towards climate variability.

Meuwissen (2019) and Meuwissen (2018) ref. [10,45] proposed a methodological framework to assess the resilience of agricultural systems using the adaptive cycle as a key concept. This framework includes three fundamental capacities or indicators for analyzing or assessing the resilience of agricultural systems: (1) robustness, (2) adaptability, and (3) transformability; and six attributes of resilience: (1) learning, (2) production, (3) networks, (4) governance, (5) diversity, and (6) resources. They recommended that methods for assessing resilience should focus on specifying or addressing the following questions:

• The limits of the system and its configuration (resilience of what?) (or specific resilience).
• The challenges of interest to the system (resilience for what?).
• The essential functions of the system (resilience for what purpose?).

Cabell (2012) [24] used the premise that agroecosystems are too complex to measure resilience accurately and, based on a thorough bibliographic review, proposed a set of thirteen indicators (without specifying the units of measure) that, when identified in an agroecosystem, suggest that it is resilient and able to adapt and transform: (1) socially self-organized, (2) ecologically self-regulated, (3) appropriately connected, (4) functional and response diversity, (5) optimally redundant, (6) spatial and temporal heterogeneity, (7) exposed to disturbance, (8) coupled with local natural capital, (9) reflective and shared learning, (10) globally autonomous and locally interdependent, (11) honors legacy, (12) builds human capital, and (13) reasonably profitable.

Peterson (2018) ref. [46] proposed an approach to operationally evaluate the resilience of agroecosystems. This proposal is an adaptation of the approaches developed by [47,48], wherein resilience is related to productive functions: “(1) productivity, (2) stability, (3) resistance to declines in yield components or growth parameters and their supporting mechanisms in the face of disturbance, and/or (4) rapid recovery to baseline functionality when conditions improve” [46] (p. 20).

A fundamental principle for assessing the resilience of agricultural systems was proposed by ref. [49], who warned that analyzing or evaluating the resilience of agroecosystems is not an exact process, given the complexity of these systems. In some cases, it is necessary to make predictions and decisions for future scenarios. This author proposed a heuristic approach for the development of methods to assess the resilience of agroecosystems that included the following: (1) adaptive cycle, (2) panarchy, (3) adaptability, and (4) transformability. “A heuristic technique is a problem-specific approach [not guaranteed to be perfect] that employs a practical method that provides sufficient accuracy for the [proposed] goals” [50] (p. 457). It is a practical approach for immediate objectives and decision-making for a precautionary principle.

Although resilience is a fundamental property for agroecosystems, autonomous adaptation and planned adaptation to fluctuations, as proposed by [51], and proposals for specific indices to assess the resilience of agroecosystems against climate variability, called specific resilience by [52,53], are scarce or critically underdeveloped [24]. Translating them into measurable entities is an ongoing challenge [46], which has slowed the construction of proposals. These difficulties or limitations result from climates that are diverse and present uncertainties. Measuring resilience is not an exact process but it provides information for decision-making in critical scenarios that can contribute to positive outcomes, for example, maintaining productivity, ensuring food security and autonomy, and conserving natural resources [27].

Therefore, analyzing the response capacity of agroecosystems to explicitly defined disturbances, such as climate variability, through the evaluation of specific resilience is strategically important for decision-making, guarantying agricultural production, conservation of ecosystems, and the social and economic well-being of communities [10,54].

The need to build tools to assess specific resilience is greater in developing countries, because changes in the climate tend to affect the poor more because of their state of economic and social vulnerability, as stated by ref. [55] (p. 6): “[where] climate change impacts tend to be regressive, falling more heavily on the poor than the rich”. Similarly, ref. [36] (p. 24) stated that: “inequality exerts disproportionate effects through three channels: (i) increased exposure of disadvantaged groups to climate hazards, (ii) increased susceptibility to damage caused by climate hazards, and (iii) decreased ability to cope with and recover from damage”.

The objective of this paper was to propose an index to evaluate the specific resilience of agroecosystems towards climate variability, called the agroecosystem resilience index (AgRI), built with the heuristic methods proposed by [49], which weigh the components of agroecosystems and include physical, biotic, social, economic, and cultural indicators. This index identifies the adjustments needed to strengthen the resilience of agroecosystem components, reduce vulnerability, and guarantee the production and conservation of resources in scenarios of climate variability. The main methodological contribution is the use of qualitative and quantitative indicators, generated from the analysis of the main attributes of agricultural systems, with the participation of experts and use of the Delphi method. This approach improves predictability and reduces uncertainty.

Section 1 presents a conceptual definition of resilience from the general and specific approaches to climate variability and a description of the main methodological contributions and indicators for evaluating it, emphasizing the need to build methods to evaluate the specific resilience to climatic disturbances. Section 2 describes a new index to assess the specific resilience of agroecosystems to climate variability, called the agroecosystem resilience index (AgRI), and the methodological route for its development. This index includes forty key indicators of physical, biotic, social, economic, and cultural categories, weighted by experts. Section 3 presents an application of AgRI, where the specific resilience to climatic variability of citrus crops in Colombian was evaluated and qualified, analyzing the profitability of traditional systems compared to those with a higher AgRI in Section 4. Finally, Section 5 discusses the scope and opportunities for improving AgRI, as well as the general conclusions of this article.

2. Methodological and Structural Approach of AgRI

To construct the AgRI, the agroecosystem was conceptualized as a social ecological system (SES), in the terms proposed by [53,56,57,58]; from this perspective, the indicators are related to social resilience and ecological resilience. The AgRI was built within the framework proposed by [10], it considers the three capacities that are crucial to understand the resilience of agricultural systems to climatic variability:

• Robustness: ability of the agroecosystem to maintain the desired level of products despite the occurrence of disturbances [54].
• Adaptability: the ability of human communities to manage resilience [49].
• Transformability: ability of the actors and components of the agroecosystem to create a functionally new system or structures of the system when the existing configuration is unsustainable [22,49].

Likewise, they include the attributes of resilience that contribute to the resilience of agricultural systems or improve resilience indicators [10,45]: (1) learning, (2) production, (3) networks, (4) governance, (5) diversity, (6) resources, and (7) profitability.

Specifically, the indicators were selected or classified in the context of the five generic principles for assessing resilience to climatic variability (Table 1, Column 4):

(1)

Diversity: functional diversity and responses to disturbance.

(2)

Modularity: internal division of the system into independent but connected modules [59].

(3)

Openness: refers to the connectivity between systems [59].

(4)

Adjustment of feedbacks or panarchy: the responses of one part of the system to changes in other parts of the system, including the component [60].

(5)

System reserves: reserve of resources (human, economic and social capital) for the SES when responding to disturbances or shocks [61]. The reserves provide redundancy and act as a buffer that compensates for losses or failures of system functions.

The AgRI considers the fundamental questions for assessing the specific resilience of agricultural systems proposed by [10,53], and each of the indicators were selected by taking into account the answers to these questions.

(1)

Resilience of what? agroecosystem resilience.

(2)

Resiliency to what? resilience to the disturbance of climate variability.

(3)

Resiliency for what purpose? maintain productivity features, profitability, human well-being, and physical and biotic sustainability.

(4)

What resilience capacities are assessed?

• Robustness: ability to withstand unanticipated stresses and shocks.
• Adaptability: ability to change the composition of inputs, production, marketing, and management in response to disturbances and maintain the functions of the agroecosystem.
• Transformability: ability to significantly change the internal structure and feedback mechanisms of the agroecosystem in response to disturbances.

(5)

What improves resilience? evaluation of the attributes that assess resilience through indicators. These attributes represent the individual and collective competencies and the enabling environment that enhances resilience capacities [10].

The methodology for calculating AgRI is made up of the following stages: (1) selection of indicators and organization into components and categories; (2) weighting indicators, components, and categories; (3) assignment of the interpretation scales of the indicators; (4) equations for the calculation of AgRI; and finally, (5) interpretation of the AgRI (Figure 1).

Figure 1. Flow chart of the AgRI methodological approach.

2.1. Selection of Indicators and Organization into Components and Categories

Forty key indicators were selected that, according to scientific evidence, are related to resilience to climatic disturbances (references are listed in Table 1, Column 7), generating a hierarchical structure that was divided into five categories made up of 13 components and 40 indicators. The five categories were: (1) physical-biotic, (2) agroecological, (3) sociocultural, (4) economic, and (5) technological (Figure 2).

Figure 2. Hierarchical structure of AgRI.

As proposed in this study, the AgRI has categories, components, and indicators, both structural and link. The structural ones are fundamental for the resilience of agroecosystems, because they interact directly with the disturbance and have an inherent resilience, determined by their intrinsic characteristics (Figure 3). These components can increase the acquired resilience and optimize the response, in the terms proposed by [62]. In this process, the link categories facilitate the interaction of structural categories, generating a synergistic relationship, and the emergence of the total resilience of agroecosystems in the face of climate variability. A specific analysis of the resilience of each indicator would allow making cultural or technological adjustment decisions, so that agroecosystems can continue functioning and communities can maintain their well-being. In this context, a specific resilience assessment can determine the response of agroecosystem components to climatic disturbances and identify priority areas for intervention [63].

Figure 3. AgRI structural and link categories.

2.2. Weighting of Categories, Components, and Indicators

Mikulic (2015) ref. [64] (p. 132) recommended weighting components and indicators of the social ecological systems (for example, agroecosystems), considering inherent differences in interactions with disturbances: “the weighting of the indicators becomes very important because not all indicators have to be equally important to explain the underlying sustainability phenomenon.” Therefore, in this article, higher weights were assigned to the indicators of greater resilience to climatic disturbances, based on the approach proposed by [65], because the weights have a significant effect on the index.

Based on these recommendations, a review of references related to indicators of resilience to climate variability (references are listed in Table 1, Column 7 and [10,11,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44]) was carried out, and a group of indicators and their respective weightings were constructed.

Subsequently, a group was formed with 30 (thirty) Colombian experts, qualified in the areas of environmental sciences, soils, crop protection, plant physiology, plant breeding, animal production, agricultural economics, rural sociology, edaphology, and agricultural administration. With the Google^® platform and the Delphi method [66], three successive rounds of surveys were carried out, in which initial weighting values were proposed (Table 1, column 5) and modified (Table 1, column 6) for each of the categories, components, and indicators, according to the criteria of the experts. The final values are presented in Table 1, column 6.

Table 1. Weighting of categories, components, and indicators.

Category Weighting	Component Weighting	Indicators	Generic Principles for Assessing Resilience	Proposed Weighting	Adjusted Weighting	References
Physico-Biotic (31.24)	Soils (18.16)	Slope %	Diversity	2.00	1.89	[67,68,69]
		Type of erosion	Diversity	2.00	2.06	[70,71,72,73]
		Soil drainage	Diversity	2.00	2.00	[74,75,76,77,78,79]
		Effective depth	Diversity	2.00	1.87	[78,80,81,82]
		Fertility	Diversity	2.00	2.30	[82,83,84,85]
		Land use	Diversity	5.00	4.91	[36,68,85,86]
		Soil conservation practices	Diversity	3.00	3.13	[86,87,88,89]
	Water (13.08)	Availability of water for irrigation	Adjustment	6.00	5.74	[69,90,91,92,93,94,95]
		Irrigation water quality	Adjustment	4.00	4.17	[69,90,91,92,93,94,95,96]
		Water conservation practices	Adjustment	3.00	3.17	[89,90,97,98]
Sociocultural (25.78)	Capacities (8.78)	Other management practices	Adjustment	3.00	2.87	[11,44,97,98,99]
		Perception- Consciousness	Adjustment	3.00	2.91	[100,101,102,103,104]
		Capacity for Action	Adjustment	3.00	3.00	[105,106,107,108]
	Public services and social security (3)	Availability of drinking water	Adjustment	1.00	1.00	[90,109,110,111]
		Energy availability	Adjustment	1.00	1.00	[112,113,114,115,116]
		Health care	Adjustment	1.00	1.00	[11,117,118,119,120]
	Housing (1)	Housing quality *	Adjustment	1.00	1.00	[121,122]
	Land structure (7.95)	Land tenure	Adjustment	4.00	4.30	[123,124,125]
		Farm size **	Diversity	4.00	3.65	[1,126,127,128]
	Competencies (5.05)	Training offer	Adjustment	2.00	1.87	[11,129,130,131]
		Level of schooling	Adjustment	1.00	0.98	[23,129,132,133,134]
		Participation in organizations	Adjustment	2.00	2.20	[135,136,137,138]
Technological (11.87)	Practice (6.87)	Agricultural practices	Adjustment	3.00	2.87	[23,139,140,141]
		Sustainable postharvest practices	Adjustment	3.00	2.83	[142,143,144,145,146,147]
		Weed Management	Adjustment	1.00	1.17	[83,148,149,150,151,152]
	Technical assistance (2)	Technical assistance availability	Adjustment	1.00	1.00	[96,153,154]
		Type of technical assistance	Adjustment	1.00	1.00	[91,153,155,156]
	Information management (3)	Management of climatic information	Adjustment	2.00	1.96	[93,154,157,158]
		Administration Record	Adjustment	1.00	1.04	[159,160]
Economic (15.98)	Financial capacity (8.99)	Savings Capacity	Reserves	2.00	2.30	[45,161]
		Productivity	Reserves	6.00	5.39	[162,163,164,165]
		Availability and access to credit	Reserves	1.00	1.30	[95,166,167,168,169]
	Market (6.99)	Destination of Production/generation of income	Adjustment	2.00	2.26	[169,170,171]
		Access to credit services	Adjustment	1.00	1.30	[172,173,174,175]
		Generation of Added Value	Reserves	4.00	3.43	[8,176]
Agroecological (15.13)	Agrobiodiversity and connectivity (15.13)	Connection with the Main Ecological Structure	Modularity	5.00	5.00	[177,178,179,180,181]
		Extension of External Connectors	Openness	2.00	1.91	[42,182]
		Extension of Internal Connectors	Openness	2.00	1.96
		Diversity of the External Connectors	Diversity	3.00	3.13	[180,182,183,184,185,186]
		Diversity of the Internal Connectors	Diversity	3.00	3.13
Total	100			100	100

* The overcrowding index was used: quotient between the # of people and # rooms in which they sleep. ** This indicator has been proposed as a new contribution for the evaluation of the resilience of agroecosystems to CV. Family agricultural unit (FAU) is the minimum extension of the farm to maintain the family and pay for the property. For Colombia, it is equivalent to 47.75 acres (1 h).

Table 1. Weighting of categories, components, and indicators.

Category Weighting	Component Weighting	Indicators	Generic Principles for Assessing Resilience	Proposed Weighting	Adjusted Weighting	References
Physico-Biotic (31.24)	Soils (18.16)	Slope %	Diversity	2.00	1.89	[67,68,69]
		Type of erosion	Diversity	2.00	2.06	[70,71,72,73]
		Soil drainage	Diversity	2.00	2.00	[74,75,76,77,78,79]
		Effective depth	Diversity	2.00	1.87	[78,80,81,82]
		Fertility	Diversity	2.00	2.30	[82,83,84,85]
		Land use	Diversity	5.00	4.91	[36,68,85,86]
		Soil conservation practices	Diversity	3.00	3.13	[86,87,88,89]
	Water (13.08)	Availability of water for irrigation	Adjustment	6.00	5.74	[69,90,91,92,93,94,95]
		Irrigation water quality	Adjustment	4.00	4.17	[69,90,91,92,93,94,95,96]
		Water conservation practices	Adjustment	3.00	3.17	[89,90,97,98]
Sociocultural (25.78)	Capacities (8.78)	Other management practices	Adjustment	3.00	2.87	[11,44,97,98,99]
		Perception- Consciousness	Adjustment	3.00	2.91	[100,101,102,103,104]
		Capacity for Action	Adjustment	3.00	3.00	[105,106,107,108]
	Public services and social security (3)	Availability of drinking water	Adjustment	1.00	1.00	[90,109,110,111]
		Energy availability	Adjustment	1.00	1.00	[112,113,114,115,116]
		Health care	Adjustment	1.00	1.00	[11,117,118,119,120]
	Housing (1)	Housing quality *	Adjustment	1.00	1.00	[121,122]
	Land structure (7.95)	Land tenure	Adjustment	4.00	4.30	[123,124,125]
		Farm size **	Diversity	4.00	3.65	[1,126,127,128]
	Competencies (5.05)	Training offer	Adjustment	2.00	1.87	[11,129,130,131]
		Level of schooling	Adjustment	1.00	0.98	[23,129,132,133,134]
		Participation in organizations	Adjustment	2.00	2.20	[135,136,137,138]
Technological (11.87)	Practice (6.87)	Agricultural practices	Adjustment	3.00	2.87	[23,139,140,141]
		Sustainable postharvest practices	Adjustment	3.00	2.83	[142,143,144,145,146,147]
		Weed Management	Adjustment	1.00	1.17	[83,148,149,150,151,152]
	Technical assistance (2)	Technical assistance availability	Adjustment	1.00	1.00	[96,153,154]
		Type of technical assistance	Adjustment	1.00	1.00	[91,153,155,156]
	Information management (3)	Management of climatic information	Adjustment	2.00	1.96	[93,154,157,158]
		Administration Record	Adjustment	1.00	1.04	[159,160]
Economic (15.98)	Financial capacity (8.99)	Savings Capacity	Reserves	2.00	2.30	[45,161]
		Productivity	Reserves	6.00	5.39	[162,163,164,165]
		Availability and access to credit	Reserves	1.00	1.30	[95,166,167,168,169]
	Market (6.99)	Destination of Production/generation of income	Adjustment	2.00	2.26	[169,170,171]
		Access to credit services	Adjustment	1.00	1.30	[172,173,174,175]
		Generation of Added Value	Reserves	4.00	3.43	[8,176]
Agroecological (15.13)	Agrobiodiversity and connectivity (15.13)	Connection with the Main Ecological Structure	Modularity	5.00	5.00	[177,178,179,180,181]
		Extension of External Connectors	Openness	2.00	1.91	[42,182]
		Extension of Internal Connectors	Openness	2.00	1.96
		Diversity of the External Connectors	Diversity	3.00	3.13	[180,182,183,184,185,186]
		Diversity of the Internal Connectors	Diversity	3.00	3.13
Total	100			100	100

* The overcrowding index was used: quotient between the # of people and # rooms in which they sleep. ** This indicator has been proposed as a new contribution for the evaluation of the resilience of agroecosystems to CV. Family agricultural unit (FAU) is the minimum extension of the farm to maintain the family and pay for the property. For Colombia, it is equivalent to 47.75 acres (1 h).

In the context of agroecosystem resilience towards climate variability, the experts gave greater specific weight to the components and indicators grouped in the structural category (68.90%). This is explained by the fundamental and irreplaceable role of soil and water in all biotic processes, especially in agricultural production, and their ability to interact with disturbances related to weather events, consistently with studies by [187,188], the approach proposed by [10], and the results reported in Colombia by [65]. Likewise, the land tenure type was highlighted, since a tenant does not have the same conservation interests as a landlord. Farm size also stood out, showing the difficulties that small farmers face when making intensive use of their farm area. Limits to crop rotation and conservation practices are relevant to the capacity to respond to the effects of climate disturbances [11].

For the components and indicators in the link category, the economic factor (15.98%) is associated with the availability of credit and the generation of economic surpluses that allow the introduction of technological improvements, increased productivity, and added value; increasing specific resilience to climatic disturbances [95]. As an expression of the agroecological category, connectivity and agrobiodiversity demonstrate the importance of improving the level of connectivity of minor agroecosystems with the surrounding landscape. Experts assigned 15.13% to this category, in recognition of the role that this category plays as a link between indicators in the agroecosystem, as proposed by [42]. For the technological category (11.87%), the management of climate information at the farm level stood out. It has been reported that technology is one of the more important components when facing climate disturbances: “In particular, technology has been recognized as one of the essential enabling elements for adapting to climate change” [89] (p. 7) (Figure 4).

Figure 4. Weighting of the AgRI categories.

2.3. Assignment of Interpretive Scales to the Indicators

Considering the results obtained in the three rounds of expert consultations, each indicator was assigned a rating according to a qualitative interpretive scale, to facilitate its interpretation using the values of 1, 3, and 5. The rating of 5 was associated with attributes of high resilience, 3 with medium resilience, and 1 with low resilience. These quantitative values were assigned to qualitatively interpret the AgRI using the methodological approach, which compares each value to the others, facilitating applications by users or groups of users (Table 2).

Table 2. Interpretive scales for indicators.

Indicators
Indicator	Resilience and Ranges of Value
	Resilience High	Resilience Medium	Resilience Low
Slope	≤4%	4–10%	≥10%
Type of erosion	Weak	Moderate	Severe
Soil drainage	Well drained	Imperfectly drained	Poorly drained
Effective depth	Deep ≥1 m	Moderate ≤0.50 m	Superficial ≤0.25 m
Fertility	High ≥20 cmol/kg	Medium 10–20 cmol/kg	Low ≤10 cmol/kg
Land use	100−75%	<75–50%	<50%
Soil conservation practices	Yes	Occasional	No
Availability of water for irrigation	Optimal	Regular	Deficient
Irrigation water quality	Some 500–1000 ppm	Moderate >1000–2000 ppm	Severe >2000–5000 ppm
Water conservation practices	Permanent	Occasional	None
Other management practices	Agroecological	Mixed	Conventional
Perception- Consciousness	Environmental perception	Mixed perception	Object perception
Capacity for action	High (Willingness and commitment no restrictions)	Medium (Willingness and commitment with restrictions)	Poor (None willingness and commitment)
Availability of drinking water	Permanent	Occasional	None
Energy availability	Permanent	Occasional	None
Health care	Permanent	Occasional	None
Housing quality *	≤2 people/room	3 people/room	>3 people/room
Land tenure	Owner	Lessee	Possessor
Farm size FAU	>1 FAU	1 FAU	<1 FAU
Training offer	Yes	Occasional	None
Level of schooling	Vocational, technical, or professional	Secondary	Primary
Participation in organizations	Yes	Occasional	None
Agricultural practices	Agriculture based on the principles of environmental sustainability	Agriculture in transition to sustainability	Agriculture based on maximizing production with exogenous material and energy subsidies
Sustainable Postharvest practices	Yes	Occasional	None
Weed management	Agroecological Management	In transition Management	Conventional Management
Technical assistance availability	Yes	Occasional	None
Type of technical Assistance	Public	Private with subsidies	Private
Management of climatic information	Yes	Occasional	None
Administrative record keeping	Yes	Occasional	None
Saving capacity	Yes	Occasional	None
Productivity	Significantly higher than the national average	Like the national average	Significantly lower than the national average
Availability and access to credit	Yes	Occasional	None
Destination of production/generation of income	National	Regional	Local
Access to credit services	Yes	Occasional	None
Generation of added value	Yes	Occasional	None
Connection with the main ecological structure	≤3 L	4–5 L	≥5 L
Extension of external connectors	>70%	31–70%	<30%
Extension of internal connectors	>70%	31–70%	<30%
Diversity of external connectors	≥3 species	2 species	1 species
Diversity of internal connectors	≥3 species	2 species	1 species

* The overcrowding index was used: quotient between the # of people and # rooms in which they sleep.

2.4. Equations for the Calculation of the Agroecosystem Resilience Index (AgRI)

AgRI is a composite function of three equations, structured to be calculated sequentially and hierarchically with the values in Table 2. Initially, the resilience of components is calculated (Equation (1)); then, the resilience of categories (Equation (2)) is calculated, followed by AgRI (Equation (3)).

$$\mathrm{CompRes}={\displaystyle \sum }_{\mathrm{i}=1}^{40}{\mathrm{IndRes}}_{\mathrm{i}}\times {\mathrm{W}}_{\mathrm{i}}$$

(1)

where CompRes = Resilience of Agroecosystem Components; IndRes = Indicators of Agroecosystem Resilience; W_i = Weighting of the Indicators of Agroecosystem Resilience.

$$\mathrm{CatRes}={\displaystyle \sum }_{\mathrm{i}=1}^{13}{\mathrm{CompRes}}_{\mathrm{i}}\times {\mathrm{W}}_{\mathrm{i}}$$

(2)

where CatRes = Resilience of Agroecosystem Categories; ComponentRes = Resilience of Agroecosystem Components; W_i = Weighting of Resilience of Agroecosystem Components.

$$\mathrm{AgRI}={\displaystyle \sum }_{\mathrm{i}=1}^5{\mathrm{CatRes}}_{\mathrm{i}}\times {\mathrm{W}}_{\mathrm{i}}$$

(3)

where AgRI = Agroecosystem Resilience Index; CatRes = Resilience of Agroecosystem Categories; W_i = Weighting of Resilience of Agroecosystem Categories.

2.5. Interpretation of the Agroecosystem Resilience Index (AgRI)

The AgRI weighted the sum of 40 parameters, generating a result of 100 units. Each result was evaluated on a scale between 1, 3, and 5; therefore, the grade was in a range between 100 and 500, which was interpreted according to the parameters in Table 3.

Table 3. Value scale for the interpretation of AgRI results.

AgRI Degree	Range	Description
Low	100–250	It is necessary to make design adjustments in the ecosystem components that have a low rating, through the implementation of cultural management practices that promote innovation and adjustment in the design of the categories, components, and parameters that present greater limitations.
Medium	251–350	The agroecosystem has categories, components, and parameters with a medium capacity for an adaptive response to disturbance, which must be promoted or at least maintained.
High	351–500	The agricultural or livestock system has categories, components, and parameters with a high capacity to respond and adapt to ecosystemic or cultural disturbances, in such a way that it can continue to function. However, this requires the continual strengthening of its components.

It is important to clarify that the AgRI is a proposal that can be adjusted in terms of the number of indicators and their weight as new research is developed; however, for this study, the AgRI was applied in the context of climate variability in citrus agroecosystems in Colombia, demonstrating its usefulness, as seen in chapter 3. It can be used in agroecosystems in other countries, because the indicators are standard and have been used in various studies. The weighting can vary according to the ecological and social conditions of territories, which is a subject for future research.

3. Case Study

An application of the proposed index is presented, to analyze its operation and scope and demonstrate its relevance. The application was carried out in the Department of Meta, in the municipalities of Villavicencio, Lejanías, Guamal, and Granada (Figure 5), where 82.21% of the planting area is concentrated, with 88.06% of the area in production and 92% of the total volume of citrus species production (orange: Citrus sinensis ‘Valencia’). In total, surveys were performed for 51 farms, covering an area of 16081.62 acres, equivalent to 30.35% of the farms in production. The approach of the survey was based on identifying the specific knowledge and the resilience of the subject/object of study, in this case the citrus growers, who in the context of the theory of social-ecological systems have a better attitude of learning and adaptation. Likewise, we analyzed the perception of the attributes (different variables) of the social and ecosystem components. Analyses of the study area have shown that extreme climate variability is affecting the citrus areas, which is why it was chosen for the application of AgRI [189]. With the collected information, a database was built, which was then subjected to multivariate statistical analysis.

Figure 5. Location of the study area.

To verify the robustness of the AgRI, the data related to the typification and classification of citrus production systems [190] were taken into account.

A survey was designed for compiling information, which was validated and adjusted in 10 community workshops, where 51 citrus growers were surveyed. In total, an area equivalent to 11.3% of the total producing area was surveyed.

In the structuring, validation, and adjustment of the survey, as well as in the presentation of the project and in the socialization of the results obtained, there was the active participation of specialists, technical assistants, researchers, marketers, producers, and local communities. The data obtained in the survey were validated by the consulting experts in three rounds, according to the protocols of the Delphi methodology.

The data were validated using specialized software (R v.9), and the data obtained from the multivariate statistical analysis were validated.

A dendrogram, defined by six (6) groups or “recommendation domains” corresponding to groups of farmers, was created. In each group or “recommendation domain”, three representative productive units/groups were selected, for a total of 18 farms (n = 18). For a better understanding, the dendrogram (Figure 6) allowed visualizing the formation of six typologies or groups of farmers with great similarity of the productive units within the group. One of these groups, group 4 (Table 4), was characterized by vertical integration between its production and marketing at the agro-industrial level. This group was made up of three productive units, and following the recommendations of the statisticians and with the aim of reducing dispersion and increasing reliability in each “recommendation domain” or group, the same number of productive units was selected, so 18 farms were selected (6 groups and in each group 3 farms).

Figure 6. Clustering dendrogram in six recommendation domains [135].

Table 4. Main attributes of the recommendation domains, in six groups of citrus agroecosystems (18 farms) located in the Department of Meta, Colombia [135].

Group	Average Area (h)	Characteristics
1	6.33	Farms with poor phytosanitary management. Peasants with a low level of education (incomplete elementary school) and associativity. Farmers without technical assistance and with limited economic income. Use of exclusively family labor. Mostly owners with more than 25 years of permanence in the region, without the ability to save. Production system based on multi-strata polyculture. Low link to the credit system due to high informality in land tenure (holders of good faith). Limited generational renewal. Originally 100% settlers from the interior of the country, mainly from the departments of Huila, Tolima, and Boyacá.
2	2.3	Farms with phytosanitary limitations. Medium level of schooling (incomplete secondary). Moderate infrastructure. Incidence of severe effects associated with climatic variability. No availability of technical assistance, little articulation with the financial system. Labor supply composed of day laborers at least three times a week. 90% are owners. High intensity land use, which does not allow the incorporation of rotation. Little afforestation and generational renewal in production systems. High permanence in the region.
3	9.6	High level of infrastructure. It was possible to evidence renewal of orange crop var. Valencia (Citrus sinensis L. Osbeck) and mandarin (Citrus reliculata Blanco) by Tangelo minneola (Citrus reticulata × Citrus paradisi), technified using the Fly Dragon dwarfing pattern, which allows increasing planting density. 60% of citrus growers are linked to some type of association. They have private technical assistance and extensive experience in citrus management. 50% have savings capacity, 40% have credit. Although they have not received training in climate information management, they relate temperature to preventive phytosanitary management techniques. 100% are owners and 10% delegated administration.
4	117.33	Highly technical agribusinesses with solid logistical, administrative, technical, and financial infrastructure. Articulation to specialized markets. Processing of climatological information and incorporation into phytosanitary management. Gallery forests that promote the connectivity of minor and major agroecosystems. No limitations of a phytosanitary nature. High productivity. Lot rotation and integration of livestock species. Development of quality certification processes with a view to offering products in specialized markets. 100% are owners and have multiple investments in other productive and service sectors based in the country’s capital.
5	4.25	Renewed crops. Young plants with a lower planting age. Presence of more than three varieties of citrus and other permanent crops such as cocoa (Theobroma cacao L.), semi-annual crops such as corn (Zea mays), and annual crops such as cassava (Manihot esculenta) and plantain (Musa × paradisiaca). Plants for self-consumption and marketing of a small volume at the village level. Low level of schooling and infrastructure. High experience in crop management, medium productivity, organization, savings, and credit availability. Farmers handle climate information. 100% are owners with 40 or more years living in the region.
6	6.79	Agrotourism without productive interest in the cultivation of citrus crops. Purpose of cultivation is only to conserve and improve the landscape. Phytosanitary management is limited exclusively to the control of weeds and management of meadows with light machinery. Very low productivity destined for the consumption of hotel guests. Plantations of very advanced age over 16 years without renovation. 20% are tenants from urban centers. 30% delegated administration.

A recommendation domain is defined as the way in which groups are structured or grouped by their characteristics or attributes of homogeneity within the group and heterogeneity outside it (“ward” distance), see Figure 6. The respective analyses were carried out from the components of these productive units. The most representative ecosystemic and cultural attributes are listed below in Table 4.

3.1. Determination of Resilience by Group

In the eighteen (18) analyzed agroecosystems, the AgRI was calculated for each of the representative farms of the six groups or recommendation domains with (Equations (1)–(3)). The results are shown in Table 5.

Table 5. AgRI in six groups of citrus agroecosystems.

Indicators	Group 1	Group 2	Group 3	Group 4	Group 5	Group 6
Slope	5.61	5.61	5.61	9.35	5.61	5.61
Types of erosion	6.13	6.13	8.86	10.22	3.41	3.41
Soil drainage	6.00	6.00	7.33	10.00	3.33	3.33
Effective depth	6.86	5.61	8.10	9.35	4.36	3.12
Fertility	6.91	6.91	8.45	6.91	6.91	5.38
Land use	14.74	14.74	18.01	24.57	11.46	14.74
Soil conservation practice	5.22	7.30	7.30	15.65	3.13	5.22
Availability of water for irrigation	9.57	5.74	21.04	28.70	17.22	9.57
Irrigation water quality	18.09	12.52	12.52	20.87	12.52	12.52
Water conservation practices	3.17	3.17	3.17	15.87	3.17	3.17
Other management practices	2.87	2.87	2.87	8.61	2.87	2.87
Perception-Consciousness	4.86	8.74	8.74	14.57	8.74	2.91
Capacity for action	3.00	9.00	15.00	15.00	9.00	3.00
Availability of drinking water	5.00	5.00	5.00	5.00	5.00	5.00
Energy availability	5.00	5.00	5.00	5.00	5.00	5.00
Health care	3.67	5.00	4.33	5.00	3.00	3.00
Housing quality	3.00	3.00	4.33	5.00	2.33	3.67
Land tenure	21.52	21.52	21.52	21.52	21.52	21.52
Farm size	3.65	3.65	8.52	18.26	8.52	3.65
Training offer	3.12	5.61	5.61	9.35	4.36	3.12
Level of schooling	1.63	2.28	1.63	4.89	2.28	2.93
Participation in organizations	5.12	2.20	5.12	10.98	2.20	3.66
Cultural practices	4.78	4.78	6.70	14.35	4.78	2.87
Sustainable postharvest practices	6.59	2.83	8.48	10.36	8.48	2.83
Weed management	1.17	1.17	3.52	5.87	3.52	1.17
Technical assistance availability	1.67	1.67	4.33	5.00	1.67	1.00
Type of technical assistance	1.67	1.00	3.00	5.00	1.00	1.00
Management of climatic information	1.96	1.96	4.57	9.78	1.96	1.96
Administration record	1.04	1.74	3.83	5.22	2.43	1.04
Saving capacity	5.38	5.38	8.45	11.52	3.84	11.52
Productivity	5.39	16.17	19.77	26.96	8.99	5.39
Availability and access to credit	5.65	4.78	5.65	6.52	3.04	4.78
Destination of production/ generation of income	8.29	11.30	11.30	11.30	9.80	2.26
Access to credit services	2.17	3.04	4.78	6.52	1.30	1.30
Generation of added value	5.72	10.30	10.30	10.30	10.30	3.43
Connection with the main ecological structure	11.67	5.00	5.00	15.00	5.00	5.00
Extension of external connectors	4.46	8.29	8.29	9.57	8.29	9.57
Extension of internal connectors	3.26	4.57	7.17	9.78	3.26	5.87
Diversity of external connectors	15.65	13.57	7.30	15.65	7.30	11.48
Diversity of internal connectors	11.48	9.39	11.48	15.65	7.30	11.48
AgRI	242.74	254.55	322.02	469.01	238.23	210.35

3.2. Interpretation of the AgRI in Six Groups of Citrus Agroecosystems

Based on Table 5, the results of the AgRI were interpreted in the 18 farms grouped into six “recommendation domains” (Figure 7).

Figure 7. Agroecosystem Resilience Index (AgRI).

Although, in the analyzed agroecosystems, the natural ecosystems were incipient or did not exist, those in group 4 implemented unconventional management practices, bringing them closer to the typology of agriculture in transition. This group obtained the highest AgRI (469.01). In this case, the parameters that most affected the results were related to the biodiversity of the farms. A broad availability of water resources was manifested in the presence of summer streams, on whose bank’s gallery forests were developed and conserved, promoting connectivity with the environment. An integration of major and minor animal species (cattle, horses, poultry, and beekeeping) was observed. The driving culture practices were significant.

These management conditions, which include the articulation of animal and plant species, promote biodiversity and represent a good strategy for increasing the resilience of agricultural systems [191,192,193]. This scenario is consistent with the evidence collected by various researchers on the higher levels of resilience of biodiverse agroecosystems, as compared to conventional ones.

The studies of [194,195] stand out as demonstrating that biodiverse agroecosystems adapted and subsequently recovered more quickly and efficiently after the passage of Hurricane Mitch through Central America than conventional production systems, which were affected more. In contrast, the farms in group 6 presented severe limitations in the environmental and ecosystem components, as indicated in Table 5 and Figure 7. The result was a low AgRI: 210.35 (red color). These farmers were more interested in the nature tourism offered by the crops, to promote their hotel activities.

The farms in group 3 reached a medium AgRI: 322.02 (yellow color), explained in part by their lower availability of water (exclusively deep wells), limited connectivity with their surroundings, and limited soil conservations practices. The small farm size limited crop rotation practices because of the intensive use of the soil by these farmers. Likewise, they had surpluses that allowed them to take advantage of the financial system, as well as the availability of private technical assistance.

The farms in group 2 were qualitatively classified with a medium AgRI of 254.55 (yellow color). In this group of farms, significant limitations in the availability and quality of water sources were also observed in all the parameters of the technological, economic, and sociocultural categories. In this regard, Colombia should guide and prioritize programs, plans, and incentives, so that citrus growers in this group have the possibility of improving their medium resilience associated with the moderate availability of economic resources.

The agricultural systems in group 4 had greater agrobiodiversity (presence of different herbaceous, shrub, and tree plant species of different height), expressed in a greater connectivity and diversity of external connectors. They stood out because of their greater institutional articulation, which was expressed in different ways: technical assistance, credit support, management of marketing channels, and the generation of added value, which allowed them to obtain greater resources that they then reinvested in technology and, ultimately, in greater competitiveness. Cultural and community resilience increases through training processes that strengthen the adaptive capacities of farmers, especially in tropical regions, where much of the world’s biodiversity is found [196,197,198].

4. Financial Evaluation of Agroecosystems Resilient to CV

Cabell (2012) ref. [24] (p. 9) stated that resilient agroecosystems must be profitable: “If agroecosystems are to continue to meet human needs, those who manage them must have their needs met as well. Farmers and farm workers should be able to make a living from work directly related to their labor, if they want to, without depending too much on off-farm income or subsidies”. The viability of productive systems with a conventional (less resilience to climate variability) and agroecological (greater resilience to climate variability) production approaches was analyzed based on economic evaluation indicators.

The cost structure was determined, including fixed costs, and the following indicators were analyzed by collecting information from primary sources and the works of [135,199]:

• Internal Rate of Return (IRR): the benefit received, expressed as a percentage of interest (%) when the investment is made, using the opportunity interest rate (TIO) ≥ 4.77% effective per year.
• Net Present Value (NPV): indicates the economic viability of the project (>0).
• Benefit/Cost Ratio (B/C): analyzes the relationship between the present value of gross income and expenses (>1).
• Investment Payback Period (PRI): the time that elapses for the investor to recover the invested capital.

The data for the analysis were grouped into three categories: (i) production costs, (ii) harvest costs, and (iii) income. The production costs were estimated using the records of the inputs and labor. These data were in turn grouped into the following categories: (i) labor, (ii) propagation material, (iii) machinery, (iv) amendments and fertilizers, (v) pesticides, (vi) equipment and tools, and (vii) other costs. Harvest costs were grouped as follows: (i) collection of fruit, (ii) internal transportation to collection centers, and (iii) post-harvest in farm. Income was calculated according to the average sale price in the market, while taking productivity (ton/ha) into account. The information on the cost structure for this type of crop was assessed over a 12-year horizon, because at that time the crop shows physiological maturity and productivity, which is expressed in a greater adaptation to environmental conditions. Table 6 shows the results of the financial analysis.

Table 6. Analysis of financial evaluation indicators for resilient agroecosystems.

	Internal Rate of Return (IRR)	Net Present Value (NPV)	Cost/Benefit Ratio (C/B)	Investment Recovery Period (IRP) in Years	Productivity ton/ha
Agroecosystems with higher AgRI	27%	73.990	1.71	4.95	23.11
Agroecosystems with lower AgRI	negative	−36.868	0.68	negative	11.84

Source: [135,199].

Internal Rate of Return IRR: The estimated TIO was 4.77% effective per year; for the agroecological production system, the IRR was 560% higher than in the conventional approach (negative); that is, this system was economically unsustainable and unable to meet the financial commitments.

Net Present Value (NPV): For the conventional system, this was negative, verifying that this system is not capable of generating additional value, causing economically significant losses to the producers.

C/B ratio: For the agroecological system, the ratio between investment and the generation of additional value for each peso invested was 1.71, showing a greater efficiency in the rational use of resources available at the local level and optimizing production costs at 251%, as compared to the conventional system, presenting the highest production costs in all analyzed components.

Investment Recovery Period (PRI): According to the income and expenditure structure, a production system with an agroecological approach should reach financial equilibrium in 4.95 years. The conventional production systems were unable to recover their investments in the analyzed time horizon, which should persuade producers to invest under the agroecological approach, which is more resilient to climate variability.

5. Discussion

This article presents a weighted composite index to evaluate specific resilience to climatic variability in agroecosystems, called the AgRI.

The AgRI reduces the gaps in resilience analysis identified by [37,47,131] and that are related to the need for tools to evaluate, in measurable entities, the specific resilience linked to climatic disturbances, applicable in non-urban environments, and that include proxies or link indicators between the ecological, social, and economic components of agroecosystems.

The AgRI was designed using the ecological social system approach, as proposed by [24,54,55,56], and it is composed of forty (40) indicators that describe key aspects for increasing or decreasing the inherent or acquired resilience of agroecosystems through attributes related to adaptability, transformability, and robustness, as proposed by [9,25,26,47], and the general principles of resilience presented by [9,167]. The AgRI key indicators are easy to use for those with high, moderate, and low levels of education or technical or professional training, which is very common in developing countries. Likewise, they were chosen for use in countries where information is limited. The SHARP methodology initiative, proposed by [134] to assess the climate resilience of farmers and herders, is a great contribution, but the main limitation is the 138 indicators that make up its structure, and where such information is not always available in developing countries. When analyzing the methodologies for evaluating the general resilience of agroecosystems consulted in the literature review and compared with the proposed method (AgRI), the following differences could be evidenced: A high number of components and synthetic variables; Simultaneous evaluation of components in different categories; Lack of or moderate weighting of the social ecological components makes analysis and applicability difficult; If system components have identical, linear response capacities, the method ignores the reality that the components of any system have differential attributes associated with their nature and composition; Lack of validation; They do not consider the specific resilience to climate variability.

However, since the AgRI is made up of a limited set of indicators, it can ignore some characteristics in agroecosystems that strengthen resilience. In this sense, the incorporation of other key indicators can be addressed in future research.

Some limitations that the AgRI may have, and that constitute topics for future research, include the following: (a) using 40 key indicators may ignore some characteristics of agroecosystems that strengthen resilience; (b) weighting indicators in AgRI with experts could generate biases, giving greater or lesser weight; however, the monitoring protocols in the Delphi method are notable, which aim to establish limits to subjectivity and reduce uncertainty.

Finally, some questions or topics for future research were raised: (a) How should the weighting of AgRI be adjusted to assess the specific resilience to extreme climatic disturbances and in social and ecological systems with limited diversity? (b) What key indicators could be incorporated into the AgRI? (c) How could one increase the participation of communities in the development of the index?

The use of weighted indicators to evaluate the specific resilience of agroecosystems to climate variability was more useful than trying to measure resilience itself. This result is consistent with [25], who stated that agricultural systems are too complex to accurately measure or assess resilience.

A weakness of weighting is the subjectivity of experts; refs. [168,169,170] warned that respondents are influenced by their values, interests, and knowledge, and could bias responses by giving greater or lesser weight to indicators. A strategy proposed by [171] is to establish limits or guidelines to subjectivity, to effectively take advantage of the knowledge of experts. The consultation method carried out in this research followed the guidelines of the “Delphi” method proposed by [64], which has been validated and widely used in scientific research [172].

The application of AgRI was consistent with the hypothesis of [25], who stated that resilience arises from the unique interaction between the farmer, the farm, and the social and economic context of the territory. Some further questions can be resolved as knowledge continues to be generated: How should the AgRI weighting be adjusted to assess specific resilience to extreme climatic disturbances and in social and ecological systems with limited diversity? What key indicators could be incorporated in the AgRI? How could one increase the participation of the communities in the development of the index? However, the relevance of the AgRI became evident in confirming that the farms that were more resilient to climate variability showed better phytosanitary status and higher productivity, and therefore presented better financial health and liquidity, a requirement of any tool to assess resilience in the terms proposed by [9].

The AgRI allowed us to observe that the farms with greater resilience to climate variability had higher productivity and greater financial performance, among other reasons, because they used ecosystem, social, and economic resources in a rational and efficient manner; this observation is consistent with the following studies:

“We define resilience of a farming system as its ability to ensure the provision of the system functions in the face of increasingly complex and accumulating economic, social, environmental and institutional shocks and stresses, through capacities of robustness, adaptability and transformability” [10] (p. 1). “If agroecosystems are to continue to meet human needs, those who manage them must have their needs met as well. Farmers and farm workers should be able to make a living from work directly related to their labor, if they want to, without depending too much on off-farm income or subsidies” ([24] (p. 9). “[We define resilience as its] Ability to maintain desired levels of agricultural outputs despite the occurrence of perturbations” [54] (p. 5).

6. Conclusions

Various approaches and methods have been proposed to assess the resilience of agroecosystems; however, they were mostly designed to assess overall resilience. This article proposed an index to evaluate the specific resilience of agroecosystems to disturbances associated with climate variability.

This index provides information to identify the components of agroecosystems to strengthen specific resilience to climate variability, to guarantee productivity. AgRI helps farming communities to make decisions about learning, adaptation, and transformation strategies related to technology, finance, and marketing, to maintain or build on acquired resilience.

This index was used to analyze the operability of the proposed model and to evaluate specific resilience in agroecosystems for citrus production in eastern Colombia, where there are extreme disturbances associated with climatic variability resulting from oscillation phenomena (ENSO). The results showed that it can be used to evaluate agroecosystems located in different territories.

The proposed index is associated with agrobiodiversity, and its application validated the hypothesis that diversified agroecosystems are more resilient and have greater financial benefits.

The agroecological production system promotes environmental sustainability and the efficient and rational management of resources.

The method proposed in this article demonstrated the ability to assess the resilience of agroecosystems. The most resilient agroecosystems maintained their productivity despite climate variability and were more profitable.