During our activities within the Biocomplexity Research Network, we have recognized that at the heart of effective research in this emerging field is the need for consilience – the synthesis of knowledge attained through linking facts and theory across disciplines. Understanding biocomplexity requires cross-disciplinary efforts (involving multiple disciplines), interdisciplinary research (examination of interactions among processes of various disciplines), and a transdisciplinary focus (a willingness to explore issues that transcend disciplinary approaches).
An excellent example of the need for this type of thinking lies in the field of population viability analysis (PVA) and its use in endangered species conservation planning. PVA methodologies typically employ computer simulation models to predict the likely response of endangered plant and animal populations to perturbations in their environment – most often wrought through human activities. These models use data on the demography of wildlife populations as input and can often become quite complex in their treatment of these biological processes. But these same models are woefully lacking in their treatment of the human activities that lead to species endangerment – in terms of both how they act today and, more importantly from the standpoint of risk assessment, how they will act differently in the future. To be more effective in species management planning, PVA methods must take account of these human processes and the ways in which they put wildlife populations at risk.
However, effective development of more inclusive models to predict species vulnerabilities is constrained by several challenges. First, the drivers of threats vary dramatically along spatial and temporal scales and across disciplinary domains (Levin 2000). Biological populations may be measured at mesoscales of tens of kilometers and their extinction risk measured in decades, viruses causing disease epidemics may operate at microscales of hours or days, and atmospheric change may operate at global macroscales across many centuries (Holling et al. 2002). The interactions of flows among physical, biological, and human domains are not as well understood as the dynamics within subcomponents of each domain (Holling and Gunderson 2002). Our understanding of coupled human-natural systems is confounded by non-linear interactions, heterogeneity of components, and flows among component parts (Levin 1999; Westley et al. 2002). Recent efforts to understand complex systems highlight the need to incorporate such complexities into predictive models, but provide fewer clear directions for how to map such complex interactions (Gunderson and Holling 2001).
Second, few existing models are robust enough to incorporate sufficiently diverse inputs. One approach to this problem has been to develop “mega-models” from the ground-up (e.g., Vanclay 1998). These highly sophisticated models are typically developed by teams of researchers to address complex but focused research problems. These models benefit from the cohesiveness that comes from development by a single team, up-front understanding of parameters and limits, and development within a standardized modeling environment. A major drawback to such models is their potential inflexibility because they are designed to address a pre-defined problem and their size and complexity restrict their utility to the original developers.
Another approach has been to tackle just one stratum of a complex problem. For example, success has been achieved through the use of global circulation models for understanding the effect of increasing carbon dioxide on atmospheric systems (Watson 2001). Such models are well suited to understand physical atmospheric changes over time or even broad impacts on ecosystems, but are less well suited to address problems at much finer scales across different domains, such as how animals in a wildlife population will respond to changing temperatures (Gitay et al. 2002).
Yet another approach has been to break down complex problems into scales and levels of complexity that are more manageable. Efforts to define species extinction risks have typically succeeded best where systems are bounded, data are widely available, and models are limited to specified biological phenomenon. PVA models are often able to satisfy many of these criteria. An important criticism of such models, however, is that they fail to consider many other biological, physical, and social processes. For example, only limited attempts have been made to integrate such analyses with models of environmental changes such as climatic changes, timber harvest, and ongoing land conversion. Moreover, single-focus models do not capture the complexity of changing interactions among biological, physical, and human systems.
We have focused our network’s research efforts on this last approach. We can construct new or utilize existing models that each focus on a specific aspect of a complex species conservation system in order to gain enhanced understanding of the drivers within those particular domains. However, we can then extend this process by devising techniques for physically linking the models into a more descriptive representation of the larger system – for example, the output of one model can serve as input to the second model, or as some sort of modifier to that input. We have called this enhanced system a “meta-model” that, through these numerical linkages, can simulate a complex system in a more streamlined fashion, and perhaps with dramatically less computational intensity than its respective “mega-model” counterpart. In the figure below, where specific human activities like road building lead to changes in wildlife population demography, we can link together one or more models of human population growth and land-use patterns with our Vortex population viability software to better predict the direct impacts of increasing household number and size on local wildlife.
Diagrammatic model of human population growth dynamics, generalized land-use patterns, and their points of impact on wildlife population biology. From Miller and Lacy (2003).
Click on graphic for larger view
We have created an epidemiological model that simulates infectious disease processes in wildlife populations and, ultimately, its role in influencing population viability. Currently titled OUTBREAK, the program will soon be a meta-model component through its linkage with Vortex. This linkage can be accomplished through an “open-data” protocol, where specification of input for each component submodel and its descriptors of the state of the system are stored in a form that is accessible to other programs. Other computer-based methods are then employed to specify and operationalize all dynamic interactions among the models.
Conceptually, however, this approach of progressive specification should facilitate even much more complex systems analyses. We are now exploring the use of geographic information systems (GIS) as a mechanism for transferring spatial data among models – at least in those cases in which the information from the various models are all spatially structured. In a similar fashion, we are investigating the creation of an interface that will allow multiple instances of Vortex to simultaneously model interacting species, as in predator-prey or competitor systems. Individual pair-wise meta-models could then be linked to create something like an individual-based population epidemiology meta-model to a spatial model of landscape change through GIS, and a rule-based model of animal dispersal.
Thus, our approach to using models to explore the interactions among the knowledge of diverse disciplines has evolved from an original plan for merged mega-models to our current concept of developing meta-models, thereby linking systems that retain their original structure and integrity. A simpler framework of “talking models” has now evolved into “talking modelers”, with the practitioners defining the larger-scale models that connect individual specializations, rather than supermodels being developed that would define – as emergent properties of their complexity – the interactions among displaced smaller-scale models. The result is that the models and understanding of each discipline are enriched by their new connections to other systems.
Click here for references.
