Methods - Michigan Breeding Bird Atlas Website

Methods 

The Second Michigan Breeding Bird Atlas 

Table of Contents 

List of Figures..................................................................................................................... 1 

Introduction......................................................................................................................... 2 

Survey Period...................................................................................................................... 2 

Grid and Sampling Design.................................................................................................. 2 

Coverage Goals................................................................................................................... 4 

Survey Procedures .............................................................................................................. 5 

Breeding Codes................................................................................................................... 6 

Data Collection ................................................................................................................... 7 

Other Data Sources ........................................................................................................... 11 

Data Processing and Analysis........................................................................................... 12 

Editing and Verification.................................................................................................... 12 

Mapping............................................................................................................................ 12 

Limitations and Biases...................................................................................................... 13 

Literature Cited................................................................................................................. 14 

Suggested Citation ............................................................................................................ 14 

List of Figures 

Figure 1: A typical township showing Atlas blocks .......................................................... 3 

Figure 2: Michigan Atlas regions ..................................................................................... 5 

Figure 3: Atlas survey checklist form................................................................................ 8 

Figure 4: Rare bird observation form................................................................................. 9 

Figure 5: Casual bird observation card ............................................................................ 10 

Figure 6: Point count data form ....................................................................................... 11 

© 2012 Kalamazoo Nature Center 

1 of 14



Introduction 

The second Michigan Breeding Bird Atlas (MBBA II) had two major goals: to document 

the current distribution of the state’s breeding birds and to provide a comparison with 

results from the first Michigan Breeding Bird Atlas (MBBA I). In order to accomplish 

the second goal it was necessary to conduct MBBA II in a manner consistent with the 

methodology used in MBBA I. A secondary goal of MBBA II was to collect abundance 

data. This was accomplished through development of a point count system throughout 

the state, and recording the number of breeding pairs or their equivalent in each block 

survey. Data on abundance have not been considered in this publication. 

The timing of this second Atlas was based on recommendations by the North American 

Ornithological Atlas Committee that Atlas projects be repeated at approximately 20-year 

intervals (McGowan and Corwin 2008). Field work for MBBA I was carried out from 

1983 to 1988 while the field work for MBBA II extended primarily from 2002 through 

2008. For both Atlases, the Kalamazoo Nature Center served as the headquarters, 

provided the Atlas coordinator, organized the volunteers, collected and analyzed the data, 

and hired additional staff. 

A statewide Atlas project requires the labor and commitment of many people. In addition 

to paid field staff, over 800 volunteers submitted bird data, 73 authors wrote species 

accounts, 37 photographers donated their pictures, and 17 individuals reviewed data or 

articles. Additionally, several staff of the Michigan Department of Natural Resources 

(MDNR) reviewed data, chapters, and species accounts for this publication or contributed 

through participation on committees. This project would not have been possible without 

the financial and logistical support of the Michigan Department of Natural Resources and 

the numerous other supporters and contributors listed in the acknowledgements. 

Additionally, advances in technology – particularly Geographic Information Systems 

(GIS) – greatly increased the efficiency of analyzing and mapping the data. 

Survey Period 

As indicated above, field work for MBBA II was conducted primarily from 2002 through 

2008. During a pilot year (2001), a small number of surveys were done to fine-tune the 

protocol, evaluate methodology, and develop field data forms. Field work was extended 

into 2008 to fill in remaining coverage gaps and achieve coverage levels comparable to 

those of MBBA I. 

Grid and Sampling Design 

As Michigan was opened to settlement, surveyors working for the General Land Office 

used the Rectangular Public Land Survey System (PLSS) to create a grid of township, 

range, and section lines across the state (Webster and Leib 2011); these survey lines are 

regulated by the U. S. Bureau of Land Management. In Michigan, private ownership of 

some lands predates the implementation of the PLSS; these lands are commonly referred 

to as the French Claims and are not included in the PLSS. Within this system a typical 

geographic township consists of 36 one-mile-square sections arranged in a 6 mile x 6 

mile square (Figure 1: A typical township showing Atlas blocks 


2 of 14



It is important to note that the boundaries of political townships do not necessarily follow 

the underlying geographical township boundaries, which can cause confusion. Partial 

townships occur along the edges of physical or political boundaries, land survey 

correction lines, and the Michigan Meridian, the state’s central north-south line. The 

Atlas block (Figure1) is the only spatial unit that is not inherent in the PLSS system. A 

block is one quarter of a township, each being comprised of nine sections of a standard 

square township. The PLSS grid, with the addition of quarter-township “blocks,” was 

used in MBBA I (McPeek and Adams 1991). 

Sections 

6 5 4 3 2 1 

7 8 9 10 11 12 

18 17 16 15 14 13 

19 20 21 22 23 24 

30 29 28 27 26 25 

31 32 33 34 35 36 

Figure 1: A typical township showing Atlas blocks 

Blocks 

2 1 

3 4 

The MDNR provided the historical GIS files that formed the basis of MBBA I’s grid. 

Advances in GIS technology allowed staff at the Kalamazoo Nature Center to extend the 

grid for MBBA II to include all land area, such as remote islands and coastal areas. The 

Great Lakes were six feet lower during the MBBA II survey period compared to lake 

levels during the MBBA I survey period (GLERL 2010) but the lake level differences 

had a minimal impact on the number of townships and blocks. The extension of the grid 

resulted in 1,927 townships and 7,192 blocks for MBBA II, an increase of 1%. The small 

difference between the two Atlases in total number of townships and blocks does not 

appear to impact comparisons of results or affect statements of trends within this 

publication. Failure to extend the grid would have affected results for species limited to 

the shores of the Great Lakes. 

The PLSS grid is useful for four reasons. First, it employs the familiar township-rangesection 

designation used in plat books and most county maps. For example, the bulk of 

the city of Lansing is located in T-04-N R-02-W. Second, many of the state’s roads 

follow section boundaries, giving observers identifiable boundaries for Atlas blocks. 

Third, the recommended grid size for bird atlases is a multiple of 5 km or “a close 

English equivalent” (Robbins and Geissler 1990); one side of a block is 3 miles, 

equivalent to 4.8 km. And fourth, using the same grid system allowed easy comparisons 

with results from the first Atlas. 

Michigan encompasses a large mass of land and water, making it impractical to survey 

every block with limited human and financial resources. Consequently priority blocks 

were established so that coverage would be adequate to accomplish the project’s goals 

and objectives. During the first Atlas, all blocks in each township in the SLP were 


3 of 14



chosen as priority blocks; in the NLP, two of the four blocks in a township were 

randomly selected; and only one block was randomly selected in the UP. These decisions 

were based primarily on the availability of volunteers in the different regions of the state. 

Because of the difficulty of achieving coverage goals in MBBA I, fewer priority blocks 

were used in MBBA II: A single priority block was chosen in each township, regardless 

of the region. To facilitate comparisons between the Atlas projects, the MBBA II priority 

blocks are a subset of those used in MBBA I. Therefore, while a priority block was 

randomly selected out of the four possibilities per township in the SLP, there were only 

two possibilities per township in the NLP and one possibility per township in the UP. A 

secondary block was selected in each township for MBBA II in roughly the same fashion. 

Observers were directed to survey the priority blocks first. If they were unable to travel 

to a more distant priority block, the secondary and non-priority blocks could be surveyed. 

Thus, data from all blocks were accepted, regardless of priority status, with the additional 

coverage expanding our knowledge of species-specific distributions. Because block 

coverage was unequal in MBBA II and MBBA I, results have been compiled by township 

to facilitate comparisons within and between Atlases. 

Coverage Goals 

A major objective of the field work portion of MBBA II was to adequately cover each 

priority block. As in MBBA I, a minimum count of 50 species was used to signify an 

adequate level of coverage (McPeek and Adams 1991). This was an internal quality 

control measure and surveys were never cut short at this number. While blocks in some 

habitats (urban centers, agricultural areas, and deep forest) might not reach 50 species, 

this measurement was useful in identifying locations in need of additional coverage. In 

addition to number of species, coverage can be described by field hours (effort) and by 

the number of records obtained. While these metrics had no defined goal, they provide a 

useful summary of coverage information for MBBA II. The standard Atlas forms are the 

casual card, the point count form, and the survey checklist form. Effort was collected on 

two forms: The point count protocol stipulated that observers spend five minutes at each 

point, so a standard 25-point survey counted as approximately two hours, while the 

survey checklist asked for hours spent in the field. Each species listed on a form was 

considered a record. 

While effort seems a very straightforward measurement, the number of hours needed to 

cover a block varies with the skill of the individual observer. How observers recorded 

and perceived their time spent in a block also varied. Surveys from volunteers which did 

not include time were assigned field hours at the rate of one hour per 30 species reported. 

Because data on hours were not available from MBBA I, no comparisons can be made in 

terms of hours of effort with MBBA II. 


4 of 14



Survey Procedures 

The state was divided into 26 regions, each with two to four counties (Figure 2). 

Regional Coordinators helped with recruiting volunteers, distributing information, and 

monitoring coverage. If a region did not have a coordinator, MBBA II staff at KNC 

organized observers for that area. 

Figure 2: Michigan Atlas regions 


5 of 14



Atlas participants were assigned blocks typically near where they lived or visited (e.g., a 

cabin up north). Some volunteers were willing to travel to do MBBA surveys. Each 

participant was given a packet with supplies and information including a participant 

handbook, other instructions, maps, data collection forms, and pertinent information 

about local species. Newsletters and workshops provided by Atlas staff and direct 

communication with regional coordinators and other participants were the main avenues 

for field work coordination and distribution of information. Participants were encouraged 

to survey priority blocks first and to visit all habitats within that block. Once the priority 

blocks were surveyed, they could visit other areas known to have unique or rare 

ecological communities. While priority blocks were emphasized, data were accepted 

from any area of the state. When a priority block was inaccessible (e.g., roadless, 

flooded, or private property) or a poor representation of the habitat within that township, 

a secondary or non-priority block could be substituted. During MBBA II, maximizing 

species counts in each block was encouraged over spending extra time to confirm 

breeding (KNC 2004). 

Paid field staff used the “blockbusting” survey method to cover more area in less time. 

Blockbusting consisted of covering all habitats in a block in a three- to five-hour period, 

with a goal of more than 50 species. Most field staff could cover two to four blocks per 

day using this method. Staff recorded the number of pairs observed in each block and 

were asked to visit each of their blocks twice at least a week apart. 

A subset of participants conducted MBBA II point counts. The goal was to obtain point 

count data for at least 25% of priority blocks. Observers were supplied with the point 

count handbook, maps, and data forms. Participants were asked to survey 25 points, 20 

on-road and five off-road (where conditions permitted). Initial maps contained 150 

points from which to choose; later maps contained fewer points, selected from the 

original 150 using a stratified random sample provided by the MDNR. Using either map, 

participants chose 25 points that resulted in uniform distribution within their block (KNC 

2005). 

Breeding Codes 

Breeding codes used in MBBA II were changed only slightly from those used in MBBA I 

(McPeek and Adams 1991) and closely follow standards set by the North American 

Ornithological Atlas Committee (Laughlin et al. 1990). The 20 criteria fall into four 

categories: observed, possible, probable, and confirmed. These codes and their 

explanations were on the back of the survey checklist forms and casual cards and in the 

Atlas Handbook. 

OBSERVED: 

O—Species observed during its breeding season but no evidence of breeding in block. 

Individual birds in unlikely breeding habitat, flying over, or out of their normal 

breeding range without any indication of breeding belong in this category. 

POSSIBLE: 

#—Species observed in suitable nesting habitat during its breeding season. 


6 of 14



X—Singing male present in suitable nesting habitat during its breeding season. 

PROBABLE: 

S—Singing male present at same location on at least two dates at least seven days apart 

or multiple (five or more) singing males on the same date during the breeding 

season. 

P—Pair observed in suitable nesting habitat during breeding season. 

T—Territorial behavior (chasing individuals of the same species). 

C—Courtship behavior or copulation. 

N—Visiting probable nest site. 

A—Agitated behavior or anxiety calls from adult. 

B—Nest building by wrens or excavation of holes by woodpeckers. 

CONFIRMED: 

NB—Nest building by all except woodpeckers and wrens. 

PE—Physiological evidence of breeding or brooding based on bird in hand. Banders or 

biologists actually handling the birds are to use this code. 

DD—Distraction display or injury feigning. 

UN—Used nests or eggshells found. Caution: These must be carefully identified. 

FL—Recently fledged young (of altricial species) incapable of sustained flight, or downy 

young (of precocial species) restricted to the natal area by dependence on adults 

or limited mobility. 

ON—Occupied nest: Adults entering or leaving a nest site in circumstances indicating 

occupied nest (includes high nests or nest holes, the contents of which cannot be 

seen) or adult incubating or brooding. 

FY—Adults with food for young (carrying food) or feeding young. Use caution as some 

species will continue to feed young for a number of weeks after leaving the nest 

and may move some distance. 

FS—Adult carrying fecal sac. 

NE—Nest with eggs. 

NY—Nest with young seen or heard. Presence of a Brown-headed Cowbird egg or 

young is confirmation for both the cowbird and the host species. 

Data Collection 

Atlas participants collected data using standard forms supplied by MBBA II staff and 

regional coordinators. Observers could also enter their sightings using a web-based form 

designed and maintained by Keith Saylor of Northern Michigan Birding; the forms at 

were available during the second half of MBBA II. 

The survey checklist form was used when observers surveyed an entire block (Error! 

Reference source not found.). Observers recorded location (down to section, if needed), 

date(s), hours, and contact information on the front panel. Every species potentially 

breeding in Michigan was included in the checklist. Species requiring additional 

documentation were marked in bold, with an asterisk and an explanatory code. For each 

species, the highest number of pairs seen, highest breeding evidence code, and date were 

recorded. The back panels listed the breeding codes and the Atlas ethics. 


7 of 14



Figure 3: Atlas survey checklist form 


8 of 14



Additional documentation was recorded on Michigan Natural Features Inventory (MNFI) 

rare bird observation form. Documentation was required for all Michigan listed species 

(including special concern) and any species that had not been confirmed during MBBA I 

(Figure 4). 

Figure 4: Rare bird observation form 


9 of 14



Casual bird observation cards were used to collect incidental observations (Error! 

Reference source not found.). Birds encountered during the course of other activities 

(e.g., driving to work, at the dog park) could be easily reported using the casual card. 

Figure 5: Casual bird observation card 

Participants who conducted MBBA II point counts used the point count data forms 

(Error! Reference source not found.). Each five-minute count period was broken into 

three-minute and two-minute intervals. Birds were recorded in four distance categories: 

within 50 m, 50-100 m, more than 100 m, and “flyover”. The latitude and longitude of 

each point, weather conditions, and the dominant habitat were also recorded. For 

inclusion in the maps, point count data from each block were compiled onto survey 

checklist forms. 


10 of 14



Figure 6: Point count data form 

Other Data Sources 

Data from other sources were included when feasible. Many organizations and 

independent biologists were conducting species-specific avian research, population 

inventories, or geographically restricted studies of birds and many were willing to share 

their data. They are acknowledged in the Atlas Participants section. Published 

information was also used when it was relevant and could be located to the block level. 

Appropriate breeding codes were assigned by MBBA II staff (e.g., courtship behavior 

was assigned at leks). These studies seldom followed MBBA II protocols and often took 

place in unique biological communities. These data greatly improved the distribution 

maps for certain species. 

In particular, the MDNR provided MNFI with funding to support the MNFI mission, 

which also furthered the goals and objectives of MBBA II. These projects included 

surveys for secretive marshland and grassland birds, including Yellow and King Rails 

and Henslow’s Sparrow, a three-year owl survey, and surveys of Northern Goshawk and 

Red-shouldered Hawk. Other surveys provided data for federally listed species (Piping 

Plover and Kirtland’s Warbler), other Michigan listed species (Trumpeter Swan, 

Common Loon, Peregrine Falcon, Caspian and Common Terns, Prairie and Yellowthroated 

Warblers) and game species (Northern Bobwhite, Ring-necked Pheasant, Ruffed 

and Sharp-tailed Grouse, Wild Turkey, and American Woodcock). Data for colonial 

nesting birds were submitted by the Great Lakes Colonial Waterbird Survey, a project 

conducted by the University of Minnesota for the U.S. Fish and Wildlife Service and 

Canadian Wildlife Service. Data on Eastern Bluebirds and other cavity-nesting species 

were obtained from NestWatch, a project of the Cornell Lab of Ornithology. Most Purple 

Martin information was obtained from the website . 

Breeding records noted in the Michigan Bird Survey, published in Michigan Birds and 


11 of 14



Natural History, were cross-referenced with Atlas data and new records were added for 

less common species. 

KNC provided additional data from a variety of related surveys including an avian 

inventory of the Four Townships area in Barry and Kalamazoo Counties, a long-term 

survey of the birds of the Fort Custer Military Training Center, Kalamazoo Breeding Bird 

Surveys, and other projects of the Conservation Stewardship Department. 

Data Processing and Analysis 

Most Atlas volunteers submitted their data forms directly to KNC, the Atlas headquarters, 

and printouts of the data were later provided to regional coordinators. A few regional 

coordinators received the data directly, checked it, and forwarded it to KNC. As 

indicated, some volunteers entered the data on-line. Data from the North American 

Breeding Bird Survey were provided by the USGS Patuxent Wildlife Research Center. 

All data received were then entered into a Filemaker Pro TM database. To allow 

integration with ESRI-compatible GIS data sets, these data were later transferred to a 

Microsoft Access TM database. 

Each winter, data were summarized by block and species, allowing MBBA II staff to set 

objectives and assign areas in need of coverage for the coming season. 

Editing and Verification 

It took about a year to edit all the collected data. The records were verified by doublechecking 

for field and data entry errors. Some species’ breeding criteria codes were 

edited for proper use and consistency. Breeding codes of some colonial nesting species 

(i.e., Double-crested Cormorant, Ring-billed and Herring Gull, Caspian and Common 

Tern) were downgraded from possible to observed since many of these birds can be nonbreeding 

wanderers. Various species-specific filters, customized to the published 

migration periods and breeding seasons, were also applied to the data to separate possible 

early or late migrants from possible breeding birds. Preliminary maps were sent to 

species account authors and others experienced with those species; these people looked 

for observations occurring outside the known range. 

Observers were required to submit written documentation to support reports of rare 

species. This documentation was sent to the Michigan Bird Records Committee (MBRC) 

for verification. The MBRC was formed in 1987 to maintain a comprehensive list of 

birds occurring in the state and to review records of species classified as “casual” or 

“accidental” based on defined criteria (Granlund 1988). MBRC decisions determined 

whether specific records of casual or accidental species were entered into the Atlas 

database. Many atlas projects take this same cooperative approach with their state 

records committees (e.g., New York; McGowan and Corwin 2008). 

Mapping 

Species occurrence and all other maps were created using ArcMap TM 9.2 from queries 

generated from the main dataset in Microsoft Access TM . Information was compiled at the 

township level for increased visibility and comparison to MBBA I. Township-level 


12 of 14



mapping also offset the differences resulting from intensive block coverage in SW 

Michigan. Using a PLSS data layer provided by MDNR, MBBA II staff built a 

township/ block layer and extended it to cover all land areas in the state. MBBA II staff 

also extended shoreline townships to full 36 square mile areas to balance the visual 

weight of the final product. The creation of new blocks was necessary in a few areas that 

were not in the original file; this included small unpopulated islands. 

Occurrence maps for each species indicated the highest breeding category recorded for 

that township: confirmed, probable, or possible. Records in the observed category were 

not mapped. 

Limitations and Biases 

Inherently, there are limitations and biases when conducting a statewide survey of birds. 

These include observer skill and experience, the amount of effort, and the number and 

distribution of observers. All of these influence the quality of the final data set, 

especially for analysis at the block level. Summarizing results by township mitigates 

some of the effects caused by observer-related inconsistencies as well as regional 

sampling differences. 

For MBBA II, 1,733 blocks were designated as priority for coverage. This was a more 

attainable number than the 4,300 attempted during MBBA I. One limiting factor was the 

number of volunteers during MBBA II, which was less than in MBBA I. As in MBBA I, 

there was an observer bias toward surveying popular birding areas and habitats and less 

effort expended surveying on private land or land that would yield a low species count. 

Industrial, agricultural, and urban areas were less likely to be surveyed while forests, 

coastlines, parks, marshes, and edge habitat drew more attention. There was also limited 

success covering some Great Lakes islands due to their remote nature. 

Most Atlas field work took place from late May to mid-July and during the early morning 

hours. Therefore, the data tend to be biased against species that nest early or late in the 

season. The data also tend to be biased against crepuscular, nocturnal, and marshland 

birds; MNFI surveys for these species groups greatly improved coverage for Atlas II. 

The inclusion of data from species-specific surveys, such as gamebird counts or raptor 

nesting studies, also biased the data. These studies were often focused on public lands 

while much of Michigan’s land is in private hands; effort may have been less 

concentrated on private land, even when permission was granted by landowners. More 

specific references to limitations and biases can be found in the individual species 

accounts. 


13 of 14



Literature Cited 

Granlund, J. 1988. Bylaws of the Michigan Bird Records Committee and the Michigan 

bird list. Jack-Pine Warbler 66: 175-180. 

Great Lakes Environmental Research Laboratory [GLERL]. 2010. Water Levels of the 

Great Lakes. NOAA/GLERL, Ann Arbor, Michigan, USA. . Accessed 7 March 2011. 

Kalamazoo Nature Center [KNC]. 2004. Michigan Breeding Bird Atlas II Handbook. 

Kalamazoo Nature Center, Kalamazoo, Michigan, USA. 

Kalamazoo Nature Center [KNC]. 2005. Michigan Breeding Bird Atlas II Point Count 

Handbook. Kalamazoo Nature Center, Kalamazoo, Michigan, USA. 

Laughlin, S.B., J.R. Carroll, and S.M. Sutcliffe. 1990. Standardized Breeding Criteria 

Codes: Recommendations for North American Breeding Bird Atlas Projects. In C.R. 

Smith, editor. Handbook for Atlasing North American Breeding Birds. North 

American Ornithological Atlas Committee. . 

Accessed 1 January 2010. 

McGowan, K.J., and K. Corwin, editors. 2008. The Second Atlas of Breeding Birds in 

New York State. Cornell University Press, Ithaca, New York, USA. 

McPeek, G.A., and R.J. Adams, Jr. 1991. The Atlas Survey. Pages 1-9 in R. Brewer, 

G.A. McPeek, and R.J. Adams, Jr., editors. The Atlas of Breeding Birds of Michigan. 

Michigan State University Press, East Lansing, Michigan, USA. 

Robbins, C.S., and P.H. Geissler. Survey Methods and Mapping Grids. In C.R. Smith, 

editor. Handbook for Atlasing North American Breeding Birds. North American 

Ornithological Atlas Committee. . 

Accessed 1 January 2010. 

Webster, G.R., and J. Leib. 2011. Living on the Grid: The US Rectangular Public Land 

Survey System and the Engineering of the American Landscape. Pages 2123-2138 in 

S.D.D. Brunn, editor. Engineering Earth: The Impacts of Megaengineering Projects. 

Springer, New York, New York, USA. 

Suggested Citation 

Chartier, A.T., J.J. Baldy, and J.M. Brenneman, editors. 2012. The Second Michigan 

Breeding Bird Atlas. Kalamazoo Nature Center, Kalamazoo, Michigan, USA. 

Accessed online at . 


14 of 14

Methods - Michigan Breeding Bird Atlas Website

Create successful ePaper yourself

Delete template?

Save as template?