Abundance Surveys

Species relative abundance maps were generated using data obtained during the years of the Atlas Project from 37 Breeding Bird Survey (BBS) routes established throughout Missouri. The BBS was organized by the United States Fish and Wildlife Service and has been conducted annually in Missouri since 1967. These surveys are conducted by road once each June. Each route consist of 50, three-minute observation stops, one-half mile apart. During stops, surveyors identify and record every individual bird detected. Survey of a BBS route begins one-half hour before sunrise and is completed in about four hours. Volunteer BBS surveyors included many Breeding Bird Atlas Project volunteers.

To obtain additional relative abundance data and provide a more thorough distribution of data collection locations, 60 Miniroute surveys were established specifically for the Atlas Project. Methodology for Miniroute surveys was similar to the BBS except that they 1) were run twice during the breeding season, 2) included only data collected after sunrise and 3) consisted of 15 rather than 50 observation stops. Miniroutes were necessarily shorter because they fit entirely within a Breeding Bird Atlas block. This allowed sightings along the route to be recorded on the Atlas field card. However, because Miniroutes were run twice, a total of 30, three-minute observations provided data for each Miniroute. The two runs were about a week apart, with stops sampled in reverse on the second run. The one-week interval enabled surveyors to consider whether individuals sighted on the second run could be coded as "territorial" on the Atlas field card.

Miniroute and BBS data from 1986-92 were converted to birds per stop, then combined and multiplied by 100. This created the unit measure used for the relative abundance maps included in the species accounts. Where inadequate abundance data were obtained, the abundance map has been omitted. Both BBS and Miniroutes are identified in Appendix B.

Data Processing and Verification

Atlasers were asked to submit field cards to state coordinators at the completion of the field season each year, regardless of whether the survey of the block had been completed. The majority of field cards were submitted as requested by September 15, the date by which the majority of breeding activity was concluded. State Coordinators and staff checked cards to verify all necessary information was included, reviewed data and occasionally contacted Atlasers for clarification.

At the start of the Atlas Project, eight data fields were digitized: topographic quadrangle name, assigned survey block number, bird species common name, bird species' AOU (American Ornithologists' Union) number, breeding code observed, date of observation, latitude and longitude. By the end of the project, several additional databases had been created to track volunteers, hours of effort, Brown-headed Cowbird host species and abundance information.

Over the years of the Atlas Project, various computers were employed, including IBM PCs, XTs and Pentium 90s using Microsoft DOS and IBM OS2 platforms. Dbase IIIC+, Word Perfect and ArcView software were used to store, analyze and compile the data. Additional software including Adobe Illustrator, ArcPlot, and Quark XPress were involved in publication of this report.

The codes and dates recorded for each species, together with the latitude/longitude coordinates for that block, were entered into a computer. Accuracy was checked by comparing print-outs with field cards. A determination was made as to whether survey of the block was complete. If additional coverage was desired, the Atlaser was sent a new field card for the following year together with print-outs of data recorded to date, suggestions of habitats to cover, expected species and other recommendations to improve coverage. Occasionally, another Atlaser was assigned to finish an incomplete block.

Limitations and Biases

The Missouri Breeding Bird Atlas Project was designed to sample one-sixth of the state's area via randomly selected survey blocks, each representing one-sixth of a 7.5-minute quadrangle. Because of this sampling procedure, and because the survey of a block was considered complete when approximately 75 percent of the block's species had been recorded, small, patchily-distributed habitats and rare, local breeding populations were likely missed entirely or under-surveyed. Rare, colonial species were especially misrepresented by the results depending on whether their few sites fell within blocks and were found.

Even within the surveyed blocks, results almost certainly do not include every species that was actually present during the Atlas Project. Although the project was conducted over a seven-year period, individual blocks were typically surveyed during only one or two years, and species that only intermittently inhabited these sites were often missed. Species requiring special efforts to locate, particularly nocturnal birds or those occupying difficult-to-survey habitats, were also under-represented. As a result, the Atlas Project counted only a small fraction of the total populations within such groups as bitterns, rails, accipiters and owls.

Blocks surveyed during a single intensive effort, such as during Block-Busting weekends, were subject to the greatest amount of bias. Because Block-Busting occurred at the average height of the avian breeding season, those species that may have been more easily detected earlier or later were likely under-represented in the results. Additionally, Block-Busters spent an average of 9.1 hours per block compared to 17.9 hours for a normally-assigned block. Block-Busting and other cursory surveys were used primarily in northeastern and southeastern Missouri and in the Ozarks. Statewide differences in hours spent per block are shown in figure 5.

Number of Species Per Block

There were also vast differences among Atlasers in effort and identification skills. This appears to have occasionally biased distributional results when an Atlaser was assigned several neighboring blocks. A bias towards lower species' counts near state borders resulted from the decision to include data from fractions of blocks bisected by the state line.

Relative abundance data generated from BBS and Miniroute surveys were most significant for more populous species. The true abundance of rarer species, or of those that are clumped in distribution, was poorly evaluated by the surveys. For this reason, abundance maps were not included in species accounts for rarer, unevenly distributed species. Even where abundance maps and narratives on abundance were included, the results may have been biased by the number of survey routes in each natural division. The Big Rivers natural Division contained only one route and so was subject to the greatest sampling error. Therefore, caution should be used when comparing species abundance among the natural divisions.

Results and Discussion

Of Missouri's 1,210 blocks, a total of 1,207 were surveyed during the Atlas Project. Atlasers documented 71,969 records of which 34.2 percent were possible, 32.8 percent were probable and 33.0 percent were confirmed breeders.

In 71.8 percent of the blocks, 50-70 species were identified. In only four blocks were 90 or more species recorded. The statewide average number of species per block was 59.6, with a high of 96 species and a low of 17 (unfinished block). The highest average number of species per block was 63.5 in the Osage Plains and 62.7 in the Ozark Border natural divisions. The Mississippi Lowlands had the lowest average at 44.8 species per block. See table 1 for species counts by natural division. Figure 6 shows the number of species per block statewide.

A total of 181 species of birds were reported during the Atlas Project. Twenty-four of these were neither expected to breed in the state nor confirmed to breed by the Atlas Project (see Appendix D). Of the remaining 167 species, all expected to breed in Missouri, seven were recorded as possible breeders, eight as probable breeders and 149 confirmed as breeders. Table 2 lists species not confirmed to breed by the Atlas Project, but which are known from other sources to have bred in Missouri (Robbins and Easterla 1992).

Only the Black-necked Stilt was for the first time confirmed to breed in Missouri as a result of the Atlas Project. In addition, as a result of a restoration effort, Peregrine Falcons bred in the state during the later Atlas Project years. Although they were not located within Atlas Blocks, their species account has been included. Two records of breeding Gadwall and one record of breeding Northern Pintail occurred outside Atlas Blocks in 1990 during the Atlas Project (Robbins and Easterla 1992).

Home|Lists of Illustrations and Tables|Preface|Acknowledgements|Introduction|The Natural Divisions of Missouri|Guide to Species Accounts|Index of Bird Accounts|Appendix A.|Appendix B.|Appendix C.|Appendix D.|Appendix E.| Literature Cited