Use of Road Drainage Structures by Wildlife in Maryland


Road drainage structures, or culverts, are constructed to channel small streams under roadways and can be used by wildlife for passage under roads.  This box culvert had a concrete substrate.

There are almost 4 million miles of paved roads in the United States and almost 70,000 lane miles of paved roads in the state of Maryland alone.  Roads can negatively impact terrestrial habitats, resulting in habitat loss, degradation of gene flow, and direct mortality of wildlife by vehicle collisions.

Road drainage structures, most commonly known as culverts, are usually constructed to channel small intermittent and perennial streams under roadways.  These culverts are also used by wildlife for passage under roads, minimizing some of the detrimental effects of roads.  Many researchers have shown that existing road culverts are used by a wide variety of wildlife, including deer, raccoons, foxes, bears, wolves, turtles, and bobcats.

One wildlife species of particular concern is deer because collisions between motor vehicles and deer cause 29,000 human injuries, over 200 human fatalities, and cost nearly $1.1 billion in vehicle repairs annually in the U.S.  Culverts that are properly sized and placed under newly constructed roads may reduce deer-vehicle collisions by allowing deer to pass under the road rather than over it. 

Goals for this project were to quantify the extent to which culverts are used by mesofauna and white-tailed deer (Odocoileus virginianus) in Maryland and to assess the characteristics of the culverts and surrounding cover types most prevalent at those sites.  Roadside deer mortality was also compared to deer use of nearby culverts to help determine their importance in reducing deer-vehicle collisions.


This arch culvert had a gravel and sand substrate.

The survey began with 265 randomly-selected culverts on paved roads maintained by Maryland State Highway Administration, which represent approximately 25% of Maryland's paved roads.  Culverts ranged from a minimum width and height of 0.61 m by 0.61 m to a maximum width and height of 4.57 m by 4.57 m.  We were able to continuously monitor 228 culverts during the 2.3 year study.  Each culvert was surveyed for two weeks on a roughly seasonal rotating basis, which resulted in at least nine sampling occasions at each site through the duration of the project.

Culverts occurred in one of three shapes: arch, box, or cylinder.  Six different culvert substrate types were identified:  silt, sand, gravel, cobble, bare corrugated steel, and concrete.  Seven categories of fencing arrangements were also documented at the culvert sites.

Individual culverts were also measured for openness, which is thought to be an important variable affected passage of large mammals through culverts.  The distance to woody vegetation cover and the percent visibility of the opening were also measured on both ends.  Water depth at the camera site was measured at each visit.  Wildlife associations with land use and land cover (LULC) types were also examined. 

Infra-red motion detecting digital game cameras were used to document culvert usage by wildlife.  In this image, an adult doe is traveling with her faw through a box culvert.

Culvert use was documented with passive infra-red motion detecting digital cameras.  Wildlife observations during site visits were also recorded.


Almost 33,000 identifiable images were recorded of wildlife over 2.3 years.  Forty species of wildlife were recorded by the camera traps and an additional 17 species were noted by direct visual observations.  Statistical analysis was limited to the 12 species that had been recorded by camera traps in 30 or more culverts:  northern raccoon, Virginia opossum, domestic cat, woodchuck, great blue heron, red fox, humans, white-tailed deer, gray squirrel, Norway rat, gray fox, and white-footed or deer mouse.

Species affiliations with culvert characteristics were analyzed statistically.  Important characteristics included culvert width and length, water depth, culvert shape, substrate type, and fencing configuration.  Culvert use by certain species also varied on a seasonal basis as well as on a regional basis. 

White-tailed deer usage of culverts smaller than that previously documented was also observed.  Deer used culverts as small as 1.42 m wide and 0.99 m tall.  Results also suggested that white-tailed deer will use longer culverts with lower openness as long as they are wide enough and tall enough to allow the deer to pass unimpeded.


This research has demonstrated that existing culverts improve habitat permeability or connectivity for a number of diverse species.  Properly sizing culverts in the future as well as using fencing effectively may greaty improve culvert use by deer and potentially mitigate deer-vehicle collisions.  Results from this research can be used to better design or retrofit culverts to improve wildlife-habitat connectivity and reduce wildlife-vehicle collisions.

This project was funded by the Maryland State Highway Adminstration. The principal investigator is Dr. J. Edward Gates.