Here is a brief, fun, picture-filled summary of the methods we used to collect our data. If you would like the full, scienced-out version of these methods, please peruse our publications: 

Grassland bird density and diversity: coming soon

Grassland bird productivity: Byers et al. 2017

Study Area: Driftless Wisconsin

Our study took place in the beautiful Driftless Area of Wisconsin. More specifically, in the Southwest Wisconsin Grassland and Stream Conservation Area (SWGSCA) - an area designated by the Wisconsin Department of Natural Resources as having "exceptional populations" of grassland bids, many prairie remnants, and concentrations of rare wildlife. This land is characterized by oak-savanna ridge tops and historically prairie (currently agricultural) valleys. This unique landscape is here in part because the area remained unglaciated during the last ice age.

Most of our study sites were southwest of Madison, and were warm season CRP (Conservation Reserve Program). Sites were selected to span a vegetation diversity gradient from monoculture switchgrass to diverse prairie restorations. 

Fire was used as a habitat management tool on our study sites, and each site was burned at least once during our study. 

Landscape Composition

When grassland birds select a territory, they don't base their choice solely on the vegetation they place their nest in. Many of these birds are area-sensitive, meaning they require a large area of grassland in order to nest. 

To determine landuse surrounding each study site, we covermapped the landscape. Researchers surveyed the land by driving through the area surrounding our sites. They recorded the ways the land was being used on a map, and later created detailed maps using ArcGIS. Buffer areas of 200 and 800 m around the site were broken into percent cover for each landuse: grass, grass + hay, woody cover, crops, developed land, and linear hedgerows. 



Every wildlife ecologist will tell you that a huge part of the data collection they do focuses on vegetation. After all, it's what provides a home for our wildlife! We sampled vegetation yearly at each site in 16 random locations.

Measuring litter depth. The Robel pole (vertical) and quadrat frame (laying on the ground) are also pictured. 

Measuring litter depth. The Robel pole (vertical) and quadrat frame (laying on the ground) are also pictured. 

We used a modified Robel pole and a quadrat frame to quantify vegetation height-density, the proportion of ground covered by litter, and the litter depth. Basically, how tall and dense was the vegetation, and how much dead plant material was on the ground? 

We also estimated the proportion of ground covered by grass (both warm and cool season), forbs*, prairie forbs, and plant species richness. 


*herbaceous flowering plant that is not a grass, sedge, or rush

Bird Density

A big part of our study was determining how many birds were using each study site. Since each site had different plants on it and therefor different vegetative structure, knowing which birds were where helped us understand each species' vegetative preferences. 


Spot-mapping was used to determine the number of territories held by singing males on each site. We focused on grassland obligate species, but recorded every species present except red-winged blackbirds (there were just too many of them!) Researchers walked transects through study sites and recorded the location of every singing male they encountered. Surveys took place between sunrise and 10am, and sites were spot-mapped ten times between mid-May and early July.


An Eastern Meadowlark sings from a fencepost. 

An Eastern Meadowlark sings from a fencepost. 

A completed spot map at the end of the field season. 

A completed spot map at the end of the field season. 

After the data was digitized in ArcGIS, polygons were drawn around clusters of three or more points considered to be associated with one male. We called each of these at territory. 

Nest Productivity

Finding Nests

Nests are like little secrets waiting to be found in the grass - even though the parent birds try their hardest to keep nests well hidden, secret and safe. Again, we focused on finding nests of grassland obligate birds, but monitored any bird nest that we found except red-winged blackbirds (there were just too many of them!)

Nests were found using four methods: 

My field crew searching for nests using the 'rope dragging' technique. 

My field crew searching for nests using the 'rope dragging' technique. 

Rope dragging: searchers walk systematically through a field dragging a rope between them. The rope disturbs the grass as it passes about 6 inches over the ground, and this flushes birds off of the nest. Additional researchers walk behind the rope watching for flushed birds. 

Systematic walking: about five searchers walk through a field in a line, spaced about a meter apart. They walk systematic transects through the field, watching for birds flushing off of the nest because of their presence. These searchers could also use sweeping sticks (small wooden or metal rods, about 0.5 inches in diameter and about 5 feet long) to disturb the grass, allowing them to search a larger area. 

Behavioral Observations: As the grass gets taller throughout the growing season, rope dragging and systematic walking become less effective. The birds are able to hear us coming from farther away, and they flush earlier. The grass is so tall that it's hard to see where bids flush from. During this time of the year, simply finding a nice spot to sit and observe the birds is more effective! Researchers looked for birds carrying nesting material, food, or fecal sacs, and tried to use their behavior to discover where nests were hidden. 

Incidental Flushing: We spent a lot of time out in the grass: checking nests, changing camera batteries (more on this later!) spot mapping, and measuring vegetation. Any time an adult bird flushed while we were doing another activity we called it an "incidental flush". 

After finding a nest, we recorded the location using GPS and marked it with a wire flag. 

Monitoring Nests

Once we found a nest, we monitored it until it either fledged one or more chicks (success!) or until the chicks died (failure).

Nests were visited every two - three days to check their status. At each visit we recorded the nest contents, the age of any chicks, and the general nest condition. We also noted adult behavior if they were near the nest, as well as signs of cowbird activity. 

All of our vegetation measurements were also carried out at finished nests. In addition to the standard measurements, we recorded nest height and the species of plant the nest was placed in. 

This is an example of a nest card for a Dickcissel nest. 

This is an example of a nest card for a Dickcissel nest. 

A perfect Upland Sandpiper nest.

A perfect Upland Sandpiper nest.

Nest Cameras

A subset of our nests were monitored using remote infrared video cameras. 

A nest camera on a Horned Lark nest

A nest camera on a Horned Lark nest

Because birds are more likely to abandon a nest while building or laying eggs, we delayed setting up video cameras until after incubation had started. Cameras were hidden in nearby vegetation, and placed a good distance away from the nest. A 25 foot cable connected the video camera with the battery and camera box. This cable was buried under litter and dead vegetation, so as not to provide a visual cue to predators. The camera boxes and batteries were covered with camoflage fabric, both to hide it from wildlife and to protect the equipment from the sun. 

Scientific Morals: Protecting the Birds We Study

Throughout this entire study, every effort was made to minimize the impact our activities had on the wildlife in our study sites.  

Dickcissel chicks, almost ready to fledge. 

Dickcissel chicks, almost ready to fledge. 

While nest searching: because we walked transects through each field weekly, our scent was spread evenly throughout the site. This reduced the chance that predators would use our scent to find nests. 

When marking nests: Some predators are really smart! There have been studies showing that predators are able to figure out the ways scientists mark nests, and use these markings to find the nests themselves. To minimize the chance of this happening, our nest marking flags were placed a full meter away from each nest, and were randomly placed to either the north or south of nests.

While checking nests: researchers approached nests from a different direction during each check. When at the nest, they did not kneel down in the grass, and minimized the amount of contact they had with the vegetation and nest. As they left the nest, researchers replaced any disturbed vegetation and continued past the nest in the direction that they had been heading- this prevented us from creating dead-end scent trails for predators to follow.