The Ecosystem of Dark Networks: A Biological Perspective

By: Dr. Matthew M. Mars , University of Arizona
Dr. Judith L. Bronstein , University of Arizona
Dr. Patricia L. Sullivan , University of North Carolina–Chapel Hill


Organizations increasingly belong to complex networks that enable them to work together in support of shared and complementary goals. To understand this trend, scholars, policy makers, and leaders regularly seek new viewpoints from which to explore the conditions and complexities associated with human networks and organizational systems. Sociologists have developed a range of analytical models for identifying actors and organizations within formal and informal systems, and for explaining the various relational ties that link these organizations together.1 Social network analysis (SNA) has been used to describe the formation of and communication patterns within and between terrorist cells, as well as to predict the outcomes of particular cell activities.2 Many questions remain, however. Organizational scientists have begun to recognize the power of biological concepts to explain the dynamics that foster and sustain linkages between actors and organizations.3 Here, we look to the field of ecosystem ecology for insights into the conditions, relational dynamics, and complexities that underpin and sustain violent non-state actor (VNSA) networks.

We are not the first to turn to biology in an attempt to understand the factors that shape and influence organized violence. For example, Darwinian evolution has been used to frame the challenges nation-states face when responding to rapidly changing strategies deployed by terrorist groups and networks.4 We do not aim here to apply a grand biological theory to the overarching problem of violence and instability. Instead, we focus on identifying and describing certain parallels between biological ecosystems and human networks. In doing so, we challenge common assumptions about the nature of militant group networks, and propose a new framework for investigating the broader implications of removing particular militant groups from violent conflict systems.

We first introduce biological ecosystems and the concept that "keystone species" exist within them. We note the diversity of outcomes that have been observed once keystone species are removed, as well as the difficulties of identifying keystones in non-disrupted ecosystems. Next, we describe how these concepts can be applied profitably to human networks (i.e., organizational ecosystems), which we extend to militant networks. Finally, we offer three possible implications of modeling violent conflict systems as ecosystems. While we do not argue for a complete parallel with biological systems, we contend that a framework that can explain patterns and processes in nature offers an exciting lens through which to view the sociological underpinnings of violent instability.

Biological Principles

Ecology is the study of the relationships between organisms and the biological and non-biological components of their natural environments. Ecologists consider natural systems to be organized in a nested structure. In a given locale, there are individual organisms, groups of organisms of the same species (populations), antagonistic or cooperative interactions among groups of species (communities), and interactions among communities and the non-biological environment (e.g., air, water, and sunlight). We refer to these latter units as biological ecosystems to distinguish them from human organizational structures, networks, and systems, which we refer to as organizational ecosystems.

Ecosystems can be characterized as follows: First, they consist of a set of "nodes" within which multiple players function and interact. Second, these nodes are linked to each other by flows of information and resources. Third, not every node is linked to every other node; links may vary in strength and can impart positive, neutral, or negative effects. Fourth, nodes grow and shrink over time; they can be lost without the ecosystem as a whole necessarily being lost. Fifth, and most importantly for our argument, no "designer" exists; nodes and links emerge from the bottom up and either persist or fail based in part on the links that emerge among them. Thus, all ecosystems are dynamic. In the next section, we argue that human networks can be treated as ecosystems when they exhibit these same five features.

In biological ecosystems, nodes are different species (i.e., each node is a collective of individuals of the same species). Biological ecosystems can contain hundreds or even thousands of species, but certain species—keystone species—play outsized roles in structuring them. Generally speaking, keystone species are those whose removal can be expected to have exceptionally strong effects on other members of the community, and hence on the functioning of the ecosystem as a whole.5

Who are the keystone actors within a biological ecosystem? Intact systems are valuable, and it clearly would be a bad idea to answer this question by removing species to see what happens. For this reason, extensive research has been devoted to developing tools that allow one to evaluate objectively the relative impact of species on their neighbors. Some of these, such as measures of "interaction strength," are based on quantifiable traits of putative keystones.6 Network analysis has also been adapted in ways that permit putative keystones to be identified.7 Although all of these approaches have drawbacks, cumulatively they may make it possible to objectively rank the relative importance of different species for maintaining ecosystem structure and function.8 In the near future, researchers may be able to identify species whose removal can be expected to have the most profound effects.


One of the clearest results from ecosystem studies is that the removal of individual species can have a surprisingly wide range of effects. For example, naturalist Robert Paine showed that the removal of a species of California coastal starfish from a section of habitat resulted in explosive growth in the population of its two prey species, a mussel and a limpet.9 Eventually, the mussel, the better competitor for space, excluded the limpet. Hence, counterintuitively, removal of a predator led to the local disappearance of a species that it ate. This starfish is a classic keystone and indeed was the first species described using that name. This same starfish does not, however always assume a keystone role within its community. Other researchers used similar experiments to show that it is only on wave-exposed rocky shores that starfish control diversity in the manner Paine described.10 More generally, a goal of ecologists is to understand how, when, and why the effects of keystone removal depend on the local setting.


Ecosystem Science and Militant Networks

Matthew Mars, Judith Bronstein, and Robert Lusch outlined the primary similarities and differences between biological and organizational ecosystems.11 Some of these are particularly relevant to debates over strategies for eliminating militant groups and disrupting terrorist networks. In this section, we briefly introduce and contextualize ecological principles that we believe have application to the counterterrorism effort.

Like biological ecosystems, organizational ecosystems exhibit a nested structure. In organizational ecosystems, the linked nodes are different groups of humans. In the context of the political violence we are addressing, the nodes consist of VNSAs, a term we use interchangeably with armed opposition groups and militant groups. Although there are important differences in the way these terms are used in various literatures, the argument we are making is generalizable to a broad class of non-state organizations that use organized violence in opposition to the established political order. As in biological ecosystems, militant group nodes are linked by flows of information and resources, typically operational intelligence and finances. Interactions between and among armed opposition groups form militant networks (communities), and interactions among these groups and the larger environment—both natural and human constructed (e.g., government institutions)—form the violent conflict ecosystem.

The fifth characteristic of ecosystems—the serendipitous nature of their emergence—is central to the approach we propose. Although most terrorism scholarship has focused on individual, group, and dyadic-level analyses, scholars have begun to apply organizational theory and SNA to study terrorist collectivities.12 To our knowledge, however, these approaches have not applied ecosystem ecology's crucial insight that biological ecosystems are emergent, as opposed to purposefully designed. This, we argue, provides a powerful representation of the nature of "dark networks."

While humans, unlike other species, have the capacity to intentionally and strategically design complex systems, many organizational ecosystems, like their biological counterparts, emerge organically. This is especially likely to be true of systems composed of illicit organizations. Individuals may deliberately create and rationally design militant groups as organizations of individuals to achieve a collective goal. Leaders of these organizations may also attempt to forge connections to like-minded groups and create larger structures intended to increase the effectiveness and resilience of a movement. But networks of extralegal VNSAs face many barriers to sustaining complex systems for collective action toward long-term goals. Unlike legal organizations, militant groups are not embedded within institutional settings that can enforce contracts; they do not benefit from the norms and institutions that allow legal organizations to overcome barriers to collective action and make credible commitments. Due to the clandestine nature of their activities, violent opposition groups also face constraints on their ability to communicate with other groups, monitor the behavior of other actors, and develop reputations for honesty or reliability. The volatility of violent conflict systems lowers the expectation of repeated interaction, and therefore reciprocity, thus obviating effects that can sometimes facilitate cooperation among self-interested actors.

Implications for Counterterrorism

Given that VNSA systems emerge and evolve as a result of myriad lower-level interactions among groups and that militant groups are primarily motivated to ensure their own survival rather than to protect and enhance the collective condition of the system, a biological ecosystem framework may have more explanatory and predictive power than SNA. Taking an approach derived from biology challenges core assumptions from the academic and practitioner literature on political violence. Below, we develop three key implications of the biological ecosystem analogy for counterterrorism.

First, VNSA ecosystems are likely to follow patterns distinct from, and more complex than, the ones we observe among social networks that comprise licit organizations. In a network approach, "theories of structural balance (or transitivity) hypothesize that only certain patterns of positive (affect) and negative (enmity) ties can exist among three nodes. Essentially, the friend of my friend is my friend, and the enemy of my enemy is my friend." 13 But a glance at the ever-shifting constellation of alliances, mergers, splits, and rivalries among militant groups in the Middle East and Central Asia quickly reveals that interactions among groups frequently violate this pattern. Militant groups may cooperate for a time, even if many of their long-term strategic objectives are incompatible, as long as all groups believe the relationship increases their own power and prestige or advances a short-term goal. Alliances dissolve and groups split as soon as severing ties with, competing with, or even preying upon a former ally becomes advantageous, regardless of shared ideologies or stated objectives. In Iraq and Syria, for example, the past decade has seen the rise and fall of at least one hundred Sunni militant groups claiming to represent the same population and pursuing closely aligned strategic objectives.

Despite their shared interests and the obvious benefits of collaboration, the Islamic State first allied itself with and then became a rival of Ansar al-Islam; al Qaeda Central broke ties with the Islamic State, whose leader had pledged a loyalty oath to Osama bin Laden in 2004; and the Nusra Front, which initially aligned itself with al Qaeda against the Islamic State, may now be abandoning its affiliation with al Qaeda.14


Second, militant group networks are likely to be more volatile and less resilient than other social networks. There has been increasing interest in, and concern about, cooperation among terrorist groups in what is sometimes referred to as a global jihadi threat. Just as in nature, however, the mere existence of a network is not sufficient evidence of a healthy, functional, and persistent ecosystem.15 Terrorism scholar Martha Crenshaw observes that "the global jihadist ‘movement' is actually extremely fractured." 16 Accordingly, the apparent momentum of the jihadist movement, as signaled by escalating acts of terror and newly forged alliances, does not necessarily indicate a strong, organized, and stable system of global terrorism. Rather, illicit networks are prone to spontaneous failures when they come under stress as a result of insurmountable collective-action problems and security vulnerabilities.17 Attempts by the leaders of VNSA groups to forecast future conditions and strategically design the larger system of connections among groups to increase the odds of a broader movement's effectiveness and survival are unlikely to be sustainable over time. We should not expect to see militant groups designing resilient networks to advance a common cause, because illicit actors cannot credibly commit to upholding bargains when component groups could benefit from unilateral defection. Analyst Chad Serena maintains that between 2003 and 2008, the Iraqi insurgency created a tremendous amount of chaos and terror but failed to achieve its primary strategic objectives—overthrowing the central government and pushing coalition forces out of the country—because, without a centralized, hierarchical leadership, discord and competition among the multitude of militant organizations that made up the insurgency limited the network's ability to coordinate activities toward a common purpose. Serena notes that "being decentralized and networked is really effective only in the short term or if minimalist organizational goals are sought." 18

Third, the ecosystem-level effects of eliminating a node within the system, especially a suspected keystone, should not be ignored. As described above, the removal of a keystone species is expected to jeopardize the persistence of a biological ecosystem. Hence, the goal in biology is to retain the putative keystones. Conflict ecosystems also have keystone actors—militant groups, state sponsors, or financiers—whose removal has stronger effects on the system than the removal of other less critical actors. The goal in this case, one might presume, would be to eliminate the putative keystones.

Keystone removal could be central to a strategy for reducing or preventing acts of organized violence. There are, in fact, multiple studies that attempt to determine whether, and under what conditions, leadership decapitation degrades or destroys terrorist groups.19 The assumption is that eliminating a violent extralegal organization will decrease violent attacks and have a net positive effect on human security in regions plagued by militant activity. As highlighted earlier, however, ecosystem scientists have discovered that keystone actors are not easy to identify a priori. In the context of conflict ecosystems, as in biological ecosystems, keystone actors may not in fact be the most public figures, the most lethal actors, or the largest groups. Moreover, the reality is that the effects of keystone elimination on a system that has emerged from interactions among a multitude of individuals, groups, governments, and their environments is largely unknown. Just as in biological ecosystems, the removal of keystone actors from violent conflict systems can result in a variety of outcomes. The United States and its allies achieved a large measure of success in disrupting and dismantling al Qaeda Central, but consequently (and unintentionally) facilitated the rise of the Islamic State. The removal of Libya's leader Muammar el Qadafi, a longtime foe of the United States and a state sponsor of terrorism against Western interests, likewise created an opening for the Islamic State to gain a foothold in Libya. The 2007 coalition surge in Iraq exacerbated organizational cleavages and severely degraded the operational capacity of key nodes—both individuals and groups—of the Iraqi insurgent network. These tactical successes limited the insurgency's ability to achieve its long-term strategic goals but also likely resulted in an escalation of random, chaotic violence—often primarily victimizing civilians—and made the enemy less predictable.20 While removing a keystone (whether a group or a specific leader) eliminates acts of violence by that actor, we have not yet developed systematic methodologies for identifying keystone actors or predicting the system-wide effects of eliminating individuals and groups.


There are many potential applications of ecosystem models, but we are particularly excited about the potential for applying principles discovered by ecologists studying the effects of species extinction to develop testable hypotheses about the effects of eliminating particular militant groups within the VNSA organizational ecosystem. There are a number of crucial questions that could be explored using this framework. In the context of a region with multiple militant groups (pursuing a variety of goals, sometimes competing and sometimes cooperating, some more directly threatening to the United States than others, some using more brutal tactics than others), what traits identify groups that play a keystone role within the broader violent conflict ecosystem? How would eliminating a particular group affect the intensity of violence within the system as a whole? What are the effects—both beneficial and detrimental—on other VNSA nodes within the system and on the system as a whole? What other qualities of the broader environment condition the consequences of eliminating an actor within the system? There has been a tendency in both academic and policy circles to focus on the effectiveness of strategies designed to disrupt and destroy militant organizations while ignoring the wider system-level effects of eliminating any particular actor within the system. But counterterrorism strategists should be concerned with the potential unintended consequences of eliminating militant groups, as removing one node from a system clearly can have a wide range of effects.

About the Author(s):

Dr. Matthew M. Mars is an assistant professor of Agricultural Leadership and Innovation at the University of Arizona.

Dr. Judith L. Bronstein is University Distinguished Professor of Ecology and Evolutionary Biology, University of Arizona.

Dr. Patricia L. Sullivan is an associate professor of Public Policy and of Peace, War, and Defense at the University of North Carolina at Chapel Hill.

  1. John Scott, Social Network Analysis: A Handbook, 3rd ed. (Thousand Oaks, Calif.: Sage Publications, 2012).go back up
  2. Stuart Koschade, "A Social Network Analysis of Jemaah Islamiyah: The Applications to Counterterrorism and Intelligence," Studies in Conflict & Terrorism 29, no. 6 (2006): 559–75; Arie Perliger and Ami Pedahzur, "Social Network Analysis in the Study of Terrorism and Political Violence," PS: Political Science & Politics 44, no. 1 (2011): 45–50; Steve Ressler, "Social Network Analysis as an Approach to Combat Terrorism: Past, Present, and Future Research," Homeland Security Affairs 2, no. 2 (July 2006): 1–10. go back up
  3. Matthew M. Mars, Judith L. Bronstein, and Robert F. Lusch, "The Value of a Metaphor: Organizations and Ecosystems," Organizational Dynamics 41, no. 4 (2012): 271–80; Miller McPherson, Lynn Smith-Lovin, and James M. Cook, "Birds of a Feather: Homophily in Social Networks," American Review of Sociology 27 (2001): 415–44. go back up
  4. Raphael D. Sagarin and Terence Taylor, eds., Natural Security: A Darwinian Approach to a Dangerous World (Los Angeles: University of California Press, 2008).go back up
  5. L. Scott Mills, Michael E. Soulé, and Daniel F. Doak, "The Keystone-Species Concept in Ecology and Conservation," BioScience 43, no. 4 (April 1993): 219–24; H. Eden W. Cottee-Jones and Robert J. Whittaker, "The Keystone Species Concept: A Critical Appraisal," Frontiers of Biogeography 4, no. 3 (2012): 117–27.go back up
  6. Mark Novak and J. Timothy Wootton, "Using Experimental Indices to Quantify the Strength of Species Interactions," Oikos 119, no. 7 (2010): 1057–63; Nicolas J. Gotelli, Werner Ulrich, and Fernando T. Maestre, "Randomization Tests for Quantifying Species Importance to Ecosystem Function," Methods in Ecology and Evolution 2, no. 6 (2011): 634–42.go back up
  7. Mariano A. Rodriguez-Cabal, M. Noelia Barrios-Garcia, Guillermo C. Amico, Marcelo A. Aizen, and Nathan J. Sanders, "Node-by-Node Disassembly of a Mutualistic Interaction Web Driven by Species Introductions," Proceedings of the National Academy of Sciences of the United States of America 110, no. 41 (8 October 2013): 16503–7.go back up
  8. For a discussion of these drawbacks, see Cottee-Jones and Whittaker, "The Keystone Species Concept." go back up
  9. Robert T. Paine, "A Note on Trophic Complexity and Community Stability," American Naturalist 103, no. 929 (January–February 1969): 91–93.go back up
  10. Bruce A. Menge, Eric L. Berlow, Carol A. Blanchette, Sergio A. Navarrete, and Sylvia B. Yamada, "The Keystone Species Concept: Variation in Interaction Strength in a Rocky Intertidal Habitat," Ecological Monographs 64, no. 3 (August 1994): 249–86.go back up
  11. Mars, Bronstein, and Lusch, "The Value of a Metaphor." go back up
  12. Victor Asal and R. Karl Rethemeyer, "The Nature of the Beast: Organizational Structures and the Lethality of Terrorist Attacks," Journal of Politics 70, no. 2 (April 2008): 437–49; Jordi Comas, Paul Shrivastava, and Eric C. Martin, "Terrorism as Formal Organization, Network, and Social Movement," Journal of Management Inquiry 24, no. 1 (2015): 47–60; Martha Crenshaw, Mapping Terrorist Organizations (Stanford, Calif.: Center for International Security and Cooperation, Stanford University, 2010); Perliger and Pedahzur, "Social Network Analysis."go back up
  13. Emilie M. Hafner-Burton, Miles Kahler, and Alexander H. Montgomery, "Network Analysis for International Relations," International Organization 63, no. 3 (2009): 567.go back up
  14. Crenshaw, Mapping Terrorist Organizations.go back up
  15. David J. Rapport, Robert Costanza, and Anthony J. McMichael, "Assessing Ecosystem Health," Trends in Ecology and Evolution 13, no. 10 (October 1998): 397–402.go back up
  16. Martha Crenshaw, "There Is No Global Jihadist ‘Movement,'" Atlantic (11 March 2015): 1: go back up
  17. Kathleen M. Carley, Ju-Sung Lee, and David Krackhardt, "Destabilizing Networks," Connections 24, no. 3 (2002): 79–92; Mette Eilstrup-Sangiovanni and Calvert Jones, "Assessing the Dangers of Illicit Networks: Why Al-Qaida May Be Less Threatening than Many Think," International Security 33, no. 2 (2008): 7–44.go back up
  18. Chad C. Serena, It Takes More than a Network: The Iraqi Insurgency and Organizational Adaptation (Stanford, Calif.: Stanford University Press, 2014), 140.go back up
  19. Patrick B. Johnston, "Does Decapitation Work? Assessing the Effectiveness of Leadership Targeting in Counterinsurgency Campaigns," International Security 36, no. 4 (2012): 47–79; Jenna Jordan, "Attacking the Leader, Missing the Mark: Why Terrorist Groups Survive Decapitation Strikes," International Security 38, no. 4 (2014): 7–38.go back up
  20. Serena, It Takes More than a Network.go back up
Average (0 Votes)
The average rating is 0.0 stars out of 5.
No comments yet. Be the first.