Department of Defense (DoD) lands provide the best available habitat for numerous threatened, endangered and at-risk species (TER-S), and many of these species are currently managed on military lands by controlled disturbances (e.g. fires) or by de novo restoration of habitat.  However, these management strategies run the risk of converting sources (where births exceed deaths) into sinks (where deaths exceed births) or of creating ecological traps - low-quality but attractive restored habitat that bleeds animals from nearby sources, threatening metapopulation viability. In addition, disturbance during and successional changes in habitat quality following management or restoration may lead local habitat patches to cycle from sink to source status and back.

Through a combination of field studies and state-of-the-art quantitative models, researchers have used three species of endangered butterflies as a model system to rigorously investigate the source-sink dynamics of species being managed on military lands. Butterflies have numerous advantages as models for source-sink dynamics, including rapid generation times and relatively limited dispersal, but they are subject to the same processes that determine source-sink dynamics of longer-lived, more vagile taxa.

Technical Approach

For two of the focal species, the research team used previous restorations and ongoing management to study temporal source-sink dynamics. For the third, initiated new restoration, allowing the researchers to examine management effects in a controlled experiment. The research team measured demography and movement at all phases of the disturbance cycle following management or restoration. They used these data to parameterize detailed spatially explicit individual-based simulation models (SEIBMs) linked to real landscapes with dynamic changes in habitat quality due to management. They also validated the general approach by comparing patterns in the focal species to general, cross-taxa, patterns.  To further generalize the results, researchers extended the approach to other TER-S insect populations to inform additional management questions.


For the focal species work, researchers found that, in most cases, habitat restoration was creating “source” habitat. In all cases, restoration had both positive and negative effects on individual vital rates, and it was necessary to integrate these effects across the life cycle to calculate the net effects of restoration. The cross-species analysis broadly validated use of correlated random walk models with edge behavior as a basis for prediction spatial population dynamics, and revealed an important empirical pattern, specifically that animals tend to have faster movement in lower-quality habitat. This pattern means that matrix and sink habitat may increase connectivity in mixed-use landscapes, even when it does not enhance population viability.


Using these field-measured vital rates, the research team developed system specific simulation models to evaluate different management scenarios. These were presented to local managers at a capstone workshop. The work also revealed some previously unknown aspects of species biology, including the importance of species interactions (mutualism, competition, and predation) in determining the source-sink status of restoration. In some cases, this understanding immediately redirected management efforts. In others (usually, cases in which our detailed mechanistic studies differed from managers a priori opinions), researchers hope that this will cause managers to think more carefully about assumptions and perhaps prioritize research to evaluate their expectations, if not immediately changing management. More generally, the case studies demonstrate (1) the importance of measuring vital rates throughout a species life cycle, in the field, in order to assess the impacts of land management, (2) a range of simple to detail-rich modeling approaches for making these assessments, and examples of when such approaches are most useful, and (3) assessment of the main impacts of widely-used restoration tools, including herbicides, fire, artificial dams, and hardwood removal.