Ideas from a Comprehensive Piping Plover Monitoring Plan (Mike Larson, USGS)

ERDC TN-DOER-E17

September 2004

TESDMS site to allow input of data from monitoring sites with different protocols for data col-

lection and to clearly identify and assign responsibilities for data management, summary, and

report production.

Ideas from a Comprehensive Piping Plover Monitoring Plan (Mike Larson, USGS):

The International Piping Plover Census provides an example of a range-wide monitoring effort

potentially applicable to the ILT. Surveys are conducted every five years, and began in 1991

(1991, 1996, 2001). Surveys are completed during 2 weeks in June, all sites selected for surveys

are known to be occupied by plovers, or contain suitable habitat, and all adults are counted

exactly once during the surveys. Benefits of the international census include the coverage of

many sites, documentation of changes in distribution, basic habitat assessment, and the genera-

tion of interest and participation from the public.

However, there is a strong need to provide scientifically defensible estimates of Piping Plover

abundance in the Great Plains. Census efforts are often inadequate for several reasons:

Most census efforts assume every bird at a site is counted (detection probability = 1),

which rarely occurs.

Counts do not adequately account for issues of scale (i.e., not all sites within a region can

be visited, and imperfect counts within a site).

Most methods do not account for issues of spatio-temporal scales (i.e., evolutionary time,

annual redistributions, and variations during the breeding season).

Other census methods exist that are better suited for monitoring Piping Plovers in the Great

Plains.

Aspects of Piping Plover Monitoring: The basic issues involved in monitoring Piping Plover

populations are common to all monitoring efforts. First, the purpose of the monitoring effort

must be clearly stated and the population to be monitored should be clearly defined. Then, deci-

sions must be made on what specific aspects of the population are to be monitored (i.e., adult

counts, productivity, or both). Sampling efforts should be designed to account for issues con-

cerning the variable detectability of individuals, and to systematically sample regions to account

for geographic variation within the populations. Numerous methods are available to account for

the spatial context of the populations (i.e., stratified, systematic, cluster, dual-frame, adaptive),

but samples should be randomized when possible, and may include areas where birds are not

present. Finally, clear protocols must be developed that will obtain all pertinent data and that are

reproducible in a scientific context.

Dealing With Variation in Detection Probability (p): Most survey techniques assume that

counts (C) equals abundance (N) of a population (i.e., N = C). But abundance is really a function

of the count and the variation in detections (N = C/p). Variation in detectabilities can be con-

trolled by using standardized protocols that can identify and control potential sources of error.

Modeling covariates within the data will provide insights into the amount of variation of detect-

ability among individuals within the population. Similarly, specific methods can be employed

that will estimate the variation of detection probabilities. These methods include:

Capture/recapture of marked birds (requires banding effort before each survey, so this is

difficult logistically; but good statistical methods are available to analyze the data).

Multiple-observer approaches (provides a good balance between effectiveness and

feasibility).