Key features surrounding wildlife crossing structures shown to influence use include vegetation, traffic volume, and wildlife behavioral response. However, these features are not likely to be static over time, particularly when we consider the lifespan of wildlife passage infrastructure. Nonetheless, most studies monitor for fewer than three years and have focused on one or two focal species, rather than community activity. Therefore, conclusions about the effectiveness of structures may not extend longer term or for different species.
When crossing structure implementation is accompanied by vegetation changes, such as mitigation, or a reduction of vegetation cover due to construction activity, the subsequent growth of that vegetation will likely impact species use of that area, thereby influencing use of the structure. In addition to vegetative habitat structure, the presence and intensity of traffic volume impacts species movements, but there has been little opportunity to observe how wildlife may interact with existing road infrastructure in the absence of vehicle activity that co-occurs there. We addressed these gaps by conducting 10 years (2009 to 2018) of summer season monitoring of the assemblage of animals at a new road, complete with a large under-bridge crossing, in an urbanizing area. During 2012 to 2013 monitoring seasons the road was closed to all traffic, providing an opportunity to determine species response to a crossing structure in the absence of vehicle activity.
Over 7,900 wildlife tracks were identified and recorded with 28 distinct species of mammal, amphibian, reptile, and bird documented using the crossing structure. ANOSIM analysis of annual wildlife detections revealed that the community of wildlife greatly changed their use of the structure over time (R=0.551, P=0.001). Individual species responses were variable with four major groups emerging with relationship to greater frequency early in the monitoring period, later in the monitoring period, during road closure periods and others with no clear pattern. Sensitivity plots of predictor variables not captured by year effect show that the location within the structure where the detection was collected is the least important and canopy cover of 5-25% was most important.
The results of this study reinforce earlier calls for more long-term monitoring with the understanding that cost is a likely barrier to implementation. Long term studies can help researchers and managers design monitoring programs to best account for variable responses over time by documenting changes in use and working to identify covariates and interactive effects that may be driving those changes. Project managers may decide to delay monitoring until vegetation communities have had time to stabilize, avoiding erroneous conclusions about structure use. We hope that the relatively few long-term data sets available can help to characterize and inform our decision making and encourage other long-term efforts.