Trend and Annual Index Information

New! See cautions regarding Alaska analysis

Population trend estimates are the simplest measure of population change over a time period. They are expressed as a percent change per year, and in these Tables they are presented for many regions and several time periods.

Here is the header and one line of data for a "typical" trend output:

Area|--------1966-2002 trends--------| |--1966-1979---||--1980-2002---|
     Trend  P    N  (  95% CI  )  R.A.  Trend   P    N  Trend   P    N  
ALB    4.9       15  -4.2  14.1   0.14    1.3         4   7.0        13

Trends are usually presented for 3 time periods: In this example list, the years are 1966 - 2002, 1966-1979, and 1980 - 2002, although these occasionally will differ in the actual output. The BBS did not actually begin in the Central United States until 1967, and in the Western United States and most of Canada the earliest routes were surveyed in 1968. This should allow you to get an idea of overall change and change for 2 subintervals.

Areas: Trends are provided for (1) states and provinces, (2) physiographic strata (abbreviated as "S" and the strata number), (3) U. S. Fish and Wildlife Service Administrative regions (e.g. RE5) , (4) BBS regions (EA, CE, and WE), (5) United States excluding Alaska, (6) Canada (within the survey area), and entire survey area (SUR). BBS regions are groups of physiographic strata. The Eastern Region (EA) is composed of strata east of the Mississippi River in the lower 48 US, but also includes strata 28 and 29 in Canada. The Central Region (CE) extends from the Mississippi River to the edge of the Rocky Mountains. The Central Region's western-most strata are 7, 53, 35, 36, 39, 38, 37, and 30, and stratum 30 also comprises it's northern edge. The Western Region includes all strata to the west of the Central and Eastern Regions. Note that Alaska, Newfoundland, Yukon, and areas clearly to the north of the survey area (see the abundance maps for this area) are not included in any regional summary. The strata regions are illustrated in a paper by Danny Bystrak (Bystrak, D. 1981. The North American Breeding Bird Survey. Studies in Avian Biology 6:34-41, Figure 1).

Trend:Estimated trend, summarized as a % change/year.

P:Because the trends are estimates, we conduct a statistical test to determine whether the trend is significantly different from 0, and results of the test are presented as p values, indicating the significance of the trend. P greater than 0.05 indicates that we cannot reject the null hypothesis that the trend is different from 0.

2/7/2005 addition: Note that "significant" (p <0.05) long-term (1966-2005) trends are now highlighted in red for declining populations and blue for increasing populations.

N:Number of survey routes in the analysis.

( 95% CI )95% confidence interval for the trend estimate. estimated as a multiplicative (constant rate) change in counts over time, with covariables to adjust for differences in observer quality. Regional trends are estimated as a weighted average of the route trends.

R. A.:Relative abundance for the species, in birds/route. This number is an approximate measure of how many birds are seen on a route in the region.

Trend estimates are summaries of population change, and do not provide information on other patterns of population change (such as cycles) over time. Consequently, data from several time periods are presented to allow observation of temporal pattern, and regional maps of trend should be consulted to evaluate regional patterns in population change.

Annual indexes of abundance can be obtained by clicking on the hotlink associated with the region (or species) name. Please note: annual indices of abundance may not exist for all species and regions. For example, we do not provide indices for regions with fewer than 5 samples (on average) for each year in the time series, or for regions where the mean abundance of the species is less than 0.1 birds/route. If the program returns an error message, the index is not available for the region and species of interest. Of course, results of any analysis should be considered in the light of possible inaccuracies due to small sample sizes and low abundances.
Addition on 5/27/2005. The raw data used in producing the graphs is now listed below the image.