Standardized North American Marsh Bird … Standardized North American Marsh Bird Monitoring...

319

Standardized North American Marsh Bird Monitoring Protocol

COURTNEY J. CONWAY1

U.S. Geological Survey, Arizona Cooperative Fish and Wildlife Research Unit, 325 Biological Sciences East,School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, 85721,USA

1Present address: U.S. Geological Survey, Idaho Cooperative Fish and Wildlife Research Unit, P.O. Box 441141, CNR Room 103E 6th & Line Streets, University of Idaho, Moscow, ID 838444, USA

E-mail: [email protected]

Abstract.—Little is known about the population status of many marsh-dependent birds in North America but recentefforts have focused on collecting more reliable information and estimates of population trends. As part of that effort, astandardized survey protocol was developed in 1999 that provided guidance for conducting marsh bird surveys through-out North America such that data would be consistent among locations. The original survey protocol has been revised toprovide greater clarification on many issues as the number of individuals using the protocol has grown. The StandardizedNorth American Marsh Bird Monitoring Protocol instructs surveyors to conduct an initial 5-minute passive point-countsurvey followed by a series of 1-minute segments during which marsh bird calls are broadcast into the marsh following astandardized approach. Surveyors are instructed to record each individual bird from the suite of 26 focal species that arepresent in their local area on separate lines of a datasheet and estimate the distance to each bird. Also, surveyors are re-quired to record whether each individual bird was detected within each 1-minute subsegment of the survey. These dataallow analysts to use several different approaches for estimating detection probability. The Standardized North AmericanMarsh Bird Monitoring Protocol provides detailed instructions that explain the field methods used to monitor marshbirds in North America. Received 26 January 2011, accepted 2 April 2011.

Key words.—bitterns, call-broadcast surveys, detection probability, marsh birds, rails, tape playback.

Waterbirds 34(3): 319-346, 2011

The amount of emergent wetland habitatin North America has declined sharply duringthe past century (Tiner 1984; Dahl 2006; Sted-man and Dahl 2008). Populations of manymarsh birds that are dependent on emergentwetlands appear to be declining (Tate 1986;Eddleman et al. 1988; Conway et al. 1994; Con-way and Sulzman 2007). Despite evidence ofpopulation declines and the need to set re-sponsible harvest limits, a monitoring pro-gram specifically designed to determine sta-tus and estimate population trends of marshbirds is lacking. The North American Breed-ing Bird Survey includes survey data on somesecretive marsh birds, but does not adequate-ly sample emergent wetlands (Bystrak 1981;Robbins et al. 1986; Gibbs and Melvin 1993;Lawler and O’Connor 2004). Marsh birds in-clude all species that primarily inhabit marsh-es (i.e. marsh-dependent species), and manyof these species are considered “inconspicu-ous” or “secretive.” Primary species of con-cern in North America include King Rail (Ral-lus elegans), Clapper Rail (Rallus longirostris),Virginia Rail (Rallus limicola), Sora (Porzanacarolina), Black Rail (Laterallus jamaicensis),Yellow Rail (Coturnicops noveboracensis), Amer-

ican Bittern (Botaurus lentiginosus), Least Bit-tern (Ixobrychus exilis), Pied-billed Grebe(Podilymbus podiceps), Limpkin (Aramusguarauna), American Coot (Fulica americana),Purple Gallinule (Porphyrula martinica) andCommon Moorhen (Gallinula chloropus). TheU.S. Fish and Wildlife Service (USFWS) hasidentified Black Rails, Yellow Rails, Limpkinsand American Bitterns as Birds of ConservationConcern because they are relatively rare andbasic information on status and trends is lack-ing in most areas (U.S. Fish and Wildlife Ser-vice 2008). Moreover, Yellow Rails, BlackRails, Clapper Rails and King Rails are four ofthe 139 “Focal” species that USFWS has givenmanagement priority because they pose spe-cial management challenges (U.S. Fish andWildlife Service 2005). Black Rails, YellowRails and Saltmarsh Sharp-tailed Sparrows(Ammodramus caudacutus) are three of the 20species on the National Audubon Society’snational ‘Watchlist’ because they are the‘most imperiled’ species (National AudubonSociety 2007). Many U.S. states consider thesespecies threatened or of special concern forsimilar reasons. King Rails are federally en-dangered in Canada, Least Bitterns are feder-

320 WATERBIRDS

ally threatened in Canada (COSEWIC 2002)and Black Rails are federally endangered inMexico (Diario Oficial de la Federacion2002).

Populations of marsh birds may be affect-ed by accumulation of environmental con-taminants in wetland substrates because theyconsume a wide variety of aquatic inverte-brates (Odom 1975; Klaas et al. 1980; Eddle-man et al. 1988; Gibbs et al. 1992; Conway1995). Marsh birds are also vulnerable to in-vasion of wetlands by many invasive plantspecies (e.g. Lythrum salicaria, hybrid Typha,Phalaris arundinacea, Phragmites, etc.) (Gibbset al. 1992; Meanley 1992). Hence, marshbirds may represent “indicator species” forassessing wetland ecosystem quality, andtheir presence can be used as one measureof the success of wetland restoration efforts(Lewis and Casagrande 1997). Marsh birdsalso have high recreational value; many ofthese species are highly sought-after by rec-reational birders. Finally, several rails arehunted in many U.S. states and Canadianprovinces yet we lack the necessary informa-tion on population trends and status uponwhich to set or adjust sustainable harvest lev-els. For these reasons, numerous federalagencies have been cooperating to monitormarsh bird populations in North Americawith the hope of gaining better knowledgeon status and distribution of these birds andimproving estimates of population trends.Continued monitoring will also allow re-source managers to evaluate whether man-agement actions or any other activities ad-versely impact wetland ecosystems. Any ac-tion that alters water levels, alters salinity, re-duces mudflat/open-water areas, altersinvertebrate communities or reduces theamount of emergent plant cover withinmarsh habitats could potentially affect habi-tat quality for marsh birds (Conway 1995).To help achieve these goals and to help en-sure that marsh bird survey data collectedthroughout North America would be collect-ed in a consistent manner, I developed astandardized marsh bird survey protocol.The Standardized North American MarshBird Monitoring Protocol (or previous ver-sions of it) has been used to collect marsh

bird survey data at locations throughoutNorth America beginning in 1999. The pro-tocol is outlined below and is also availableon the internet (see website at http://www.cals.arizona.edu/research/azfwru/Na-tionalMarshBird/).

Objectives

The Standardized North AmericanMarsh Bird Monitoring Protocol is intendedto provide guidance to individuals planningto survey marsh birds to address different ob-jectives. The most commonly-stated objec-tives include: 1) document presence or dis-tribution of marsh birds within a defined ar-ea, 2) estimate or compare density of secre-tive marsh birds among management units,wetlands or regions, 3) estimate populationtrend for marsh birds at local or regionalscale, 4) evaluate effects of management ac-tions (often actions that target other spe-cies) on secretive marsh birds, and 5) docu-ment habitat types or wetland conditionsthat influence abundance or occupancy ofmarsh birds. The Standardized North Amer-ican Marsh Bird Monitoring Protocol allowsdata sharing and comparisons among sites.U.S. Fish and Wildlife Service has a vested in-terest in marsh bird populations and theirhabitats because marsh birds are a trust spe-cies, under the protection of the USFWS.The National Wildlife Refuge System of theUSFWS has participated in conducting stan-dardized marsh bird surveys based on thisprotocol since its inception because the ref-uge system has a disproportionate amount ofwetland within their boundaries, and themanagement actions employed by refugeshave the potential to dramatically affectmarsh bird populations. However, the proto-col has also been used (and continues to beused) by biologists in a wide variety of gov-ernmental and nongovernmental agenciesand academic institutions.

Density, Abundance and Detection Probabili-ty. Abundance is the total number of birdswithin a defined area of interest. Density isabundance divided by area; for example, thenumber of birds per hectare of wetland (orbirds/ha of emergent vegetation within a

MARSH BIRD MONITORING PROTOCOL 321

wetland). Surveys rarely count all individualspresent in the sampling area because detec-tion probability is typically less than 100%.Estimates of abundance or density rely uponestimates of detection probability and either1) a consistent positive correlation betweennumber of individuals detected during a sur-vey and number of individuals actuallypresent in the area sampled (i.e. low spatialand temporal variation in detection proba-bility), or 2) incorporating environmentalcovariates into the estimation process that ef-fectively adjust for most of the variation indetection probability. Few reliable estimatesof detection probability during marsh birdsurveys are currently available (but see Con-way et al. 1993; Legare et al. 1999; Conwayand Gibbs 2001; Bogner and Baldassarre2002; Nadeau et al. 2008). However, thesesurvey protocols incorporate methods for es-timating several components of detectionprobability (see Conway et al. 2010 for an ex-ample of how estimates of detection proba-bility derived from these methods can beuseful). Some authors have expressed skepti-cism about the value of incorporating meth-ods intended to estimate detection probabil-ity into surveys (Johnson 2008), but othershave advocated for such methods (Burham1981; Thompson et al. 1998; Thompson2002; Rosenstock et al. 2002). Focal marshbird species were those identified by a groupof marsh bird biologists as species for whichwe lack quality information on status or pop-ulation trends (Ribic et al. 1999).

Population Trend. Population trend is of-ten not well-defined, but a common defini-tion is the percent annual change in popula-tion size for a particular species at some de-fined spatial scale over some defined timeperiod. Estimates of population trend allowmanagers to determine whether local or re-gional marsh bird populations are decliningand the rapidity at which they are declining.Managers can establish a priori populationtrend thresholds or trigger points belowwhich immediate management actionshould be taken. Such actions can preventlocal extinctions by identifying populationproblems before they become severe. TheNorth American Breeding Bird Survey pro-

vides estimates of population trends forsome species of marsh birds, but has insuffi-cient data to estimate trends of the more se-cretive or rare species.

Survey Routes

The number of survey points to includewithin a state, local refuge or managementarea (or the size of the survey area selected)is often dictated by personnel time availableand other logistical constraints. A surveyroute is a permanent grouping of points thatare surveyed together by the same surveyoron the same date during a morning (orevening) survey. Each survey point shouldbelong to one (and only one) permanentsurvey route. The number of points per sur-vey route can vary among routes based onthe number of points that one surveyor cansurvey in a morning (or evening) survey win-dow (see section below titled Time of day forsurveys for a definition of ‘survey window’). Asurveyor may only be able to survey a smallnumber of points (e.g. six or eight) in amorning or evening if points are far apart.These 6-8 points would constitute a “surveyroute”. If travel between adjacent points isrelatively easy, a surveyor may be able to com-plete 15 or more points in one morning orevening survey session and hence have 15points on that survey route. All the surveypoints that make up one survey route do notnecessarily have to be associated with thesame patch of marsh. Including fewer pointsper survey route and surveying an additionalmorning or evening (rather than fewerroutes with lots of points) will typically resultin more detections (but will require addi-tional survey days) because marsh birds aretypically most vocal in the two hours sur-rounding sunrise and the two hours sur-rounding sunset (Conway et al. 2004).

Once you choose the direction withwhich you conduct a particular survey route,be consistent (e.g. always survey the pointsalong route #1 in descending order: point 12is surveyed first and point 1 is surveyed last).Being consistent in this respect will assurethat each survey point is completed at ap-proximately the same time of day during

322 WATERBIRDS

each replicate survey. Consistency in thechronological order in which points on aroute are surveyed will help to reduce thesampling variation created by diurnal de-creases in vocalization probability of marshbirds as the morning (or evening) progress-es (Conway et al. 2004). Even though metricsto estimate variation in detection probabilityare incorporated into the survey protocol,any effort to minimize sampling variation isstill advantageous.

Location of Survey Points

Fixed, permanent survey points are cho-sen and marked with inconspicuous markersin the field. Each survey point receives aunique identification number. Record thelatitude and longitude of each survey pointusing a GPS receiver. If possible, locations ofall survey points should also be plotted onmaps of each wetland. Maps should includethe direction in which the speakers arepointed during the survey at each point.Which direction to orient the speakers is notalways obvious to someone who has not sur-veyed the route before, and may create un-wanted variation in numbers detected ifspeaker direction is not consistent. Surveypoints are located on either the upland-emergent vegetation interface or the openwater-emergent vegetation interface. Con-ducting surveys at points where observersstand within contiguous patches of emer-gent marsh vegetation may not be practicalin many inland wetlands because of the dis-turbance to emergent plants (and to callingrates of marsh birds) caused by walkingthrough the dense vegetation. However, con-ducting surveys from upland edges, roadsideedges, and open water edges may createsome bias in estimation of populationtrends. In order to determine the extent towhich the placement of points biases results,surveyors should record the local context foreach survey point:

1) along a ditch, dike or berm with emer-gent vegetation on both sides,

2) along a ditch, dike or berm with emer-gent vegetation on only one side,

3) along a public road with emergent vege-tation on both sides,

4) along a public road with emergent vege-tation on only one side,

5) along a grassland/emergent edge, 6) along a scrub-shrub/emergent edge,7) along a forest/emergent edge, 8) along an open water/emergent edge,9) within a narrow water channel or tidal

creek with emergent vegetation on bothsides,

10) within a contiguous patch of emergentvegetation (also record distance fromedge), or

11) other (and provide description of pointplacement).

A point is considered “along a publicroad” if the surveyor is within 25m of theroadside during the survey. Surveyors shouldalso record the type of road (gravel, dirt,paved, etc.). Select the choice that best de-scribes the placement of the point. These da-ta are meant to provide analysts with a meth-od by which they can evaluate whether pop-ulation trend or density estimates vary de-pending on where survey points are located.If estimates vary in this regard, analysts havethe option of stratifying their estimates totake variation in the placement of points in-to account.

Point Spacing

Point spacing in previous studies has var-ied from 40m to 800m (Conway and Gibbs2001). For setting up new routes associatedwith the Standardized North AmericanMarsh Bird Monitoring Protocol, we recom-mend 400m between adjacent survey pointsto increase the total area covered by moni-toring efforts. If points are too close together(i.e. <400m apart), then the call-broadcast atone point may affect the distribution of birdsat adjacent points (because birds within ear-shot often approach call-broadcast; Conwayand Gibbs 2011) and hence cause biases inmany analyses. Surveyors who want closerpoint spacing for some local reason shouldspace points by an interval that is easily divis-ible by 400m (i.e. 200m, 100m). Analysts


would then at least have the option of usingdata only from a subset of points (those thatare 400m apart) at that particular site for theshared (pooled) data set if they choose to doso. However, birds are much more likely tohave heard the calls broadcast at prior points(and hence alter their vocal behavior) whenpoints are spaced closer than 400m apart. Inareas where survey routes have already beenestablished and surveyed in past years, retainthe original point spacing; do not delete, ig-nore or move existing survey points even ifspacing between adjacent points is very dif-ferent than 400m. In marshlands that haveaccess throughout the marsh, points shouldbe in a 400m grid system (hence, one pointper 16 ha of marsh). If not all possible pointsin the grid system can be surveyed, a randomor systematic selection of points that can besurveyed should be selected from the poten-tial survey points. Placement of survey pointswithin the wetland is a sampling design issueand observers or survey coordinators shouldconsult a statistician. In many locations,emergent marsh occurs in small patches lessthan 16 ha in size. In cases like these, includeat least one survey point at all marshes >0.5ha within the management area. Additionalsurvey points should be added at smallmarsh patches as long as they are 400m awayfrom all other survey points.

What if Area around an Existing Point is no longer Suitable Marsh Bird Habitat?

Original survey points are never droppedfrom the survey and are always visited in sub-sequent years. If no suitable habitat ispresent at an existing survey point during aparticular year (e.g. due to drought orchange in water flow), then the surveyorsshould still make an entry for that point onthe datasheet and in the database but writein the Comments column “no survey conduct-ed because suitable emergent vegetation isnot present”. Although some of the focalspecies will very occasionally use wetlandswith open water that lack emergent vegeta-tion, these species are all much more abun-dant in wetlands with emergent vegetationduring the breeding season and so surveyors

need not conduct surveys at points that nolonger have any emergent vegetation. If sur-veyors do not conduct a survey at one ormore existing points, they must record in thedatabase the reason why a survey was notconducted at those points:

1) lack of suitable habitat (due to temporarychange such as flooding, drought, mow-ing, etc.),

2) lack of suitable habitat (due to perma-nent change),

3) survey not attempted due to logisticalreasons.

Time of Day for Surveys

Surveys can either be conducted in themorning or evening. However, once one ofthese two time periods is chosen, that timeperiod for those points along the surveyroute cannot be changed. The choice ofmorning or evening survey period (and thelength of each period) should correspondwith when marsh birds are most vocal in yourarea. Vocalization probability is typicallyhighest in the two hours surrounding sun-rise and the two hours surrounding sunset(Conway et al. 2004). Choose the optimaldaily survey time(s) for your region and usethem each year. Including both morningand evening surveys into a standardizedmonitoring protocol provides added flexibil-ity and more potential survey hours for fieldpersonnel.

Morning surveys begin 30 minutes be-fore sunrise and should be completed priorto the time when marsh birds cease calling(this varies regionally, but is often two hoursafter sunrise in southern latitudes and threehours after sunrise in northern latitudes(Conway et al. 2004; Gibbs and Melvin 1993).The time in the morning when marsh birdscease calling also varies with temperatureand time of year (Conway and Gibbs 2001,2011).

Evening surveys should begin two hoursbefore sunset and must be completed by 30minutes after sunset. When conductingevening surveys, surveyors should start theirsurvey route such that they finish the last

324 WATERBIRDS

point at the time when darkness precludesthem from seeing their datasheet. The halfhour between sunset and complete darknessis often the time when detection probabilityis highest (C. Conway, unpubl. data).

Number of Surveys per Year and Seasonal Timing of Surveys

Optimal seasonal timing for surveys willvary regionally depending on breeding chro-nology of the focal marsh birds (Appendix1) in your area. Conduct at least three sur-veys annually during the presumed peak ofthe marsh bird breeding season. The peakbreeding season in each location will varyamong the coexisting marsh birds in that ar-ea. For example, American Bitterns oftenbreed earlier than both Least Bitterns andrails in some regions, and Clapper Rails andKing Rails breed earlier than Virginia Railsand Soras in some regions (also see Rehmand Baldassarre 2007). To account for varia-tion in breeding phenology (and hence vari-ation in optimal survey timing) among coex-isting species, at least one survey should beconducted within each of the three 15-daysurvey windows. The three survey windowsvary regionally and are based on the averageminimum temperatures in May in each loca-tion (Appendix 2). The three survey win-dows increase probability of conducting atleast one survey during the seasonal peak invocalization probability for all focal marshbird species in the area. In many areas, mi-grants are still moving through when thebreeding season is well underway for localbreeders. Hence, some surveys will occur pri-or to when migration is completed for manymarsh birds.

A common goal of marsh bird surveys isto estimate trends over time in the numberof breeding adults of each species, so it is op-timal to complete all three annual surveysprior to the initiation of juvenile vocaliza-tions. At least three surveys are needed toconfirm seasonal presence or absence ofsome marsh bird species in a wetland with90% certainty (Gibbs and Melvin 1993).Three replicate surveys per year are also war-ranted because personnel organizing surveys

often do not know the local timing of thebreeding cycle of the various marsh bird spe-cies at the outset of their survey effort (Rehmand Baldassarre 2007). Finally, includingthree or more surveys per year will allow forestimation of the proportion of survey routesoccupied by each species (MacKenzie et al.2002). However, if for some reason it is notpossible to conduct a minimum of three sur-veys on the area, data collected can still beused for some purposes.

Contact the program coordinator (seecontact information below) if the three an-nual survey windows do not adequately cap-ture the peak breeding seasons of the marshbird species in your area.

Surveys in Tidal Marshes

When possible, surveys in tidal marshesshould always be conducted at a similar tidalstage for each replicate survey both withinand across years. The tidal stage withinwhich to conduct local marsh bird surveysshould be based on when highest numbersof marsh birds are likely to be detected inyour area; optimal tidal stage for surveys mayvary among regions. Many salt marsh passe-rines are forced to renest during the peakspring high tide, and detection probability ishighest during the week after a high springtide. Clapper Rail surveys conducted since1972 have been timed to coincide with ahigh tide at San Francisco Bay National Wild-life Refuge, but high tide was a period of re-duced vocalization probability for ClapperRails in southern California (Zembal andMassey 1987) and for Black Rails in northernCalifornia (Spear et al. 1999). In Mississippi,detection probability and tidal height werepositively correlated for Clapper Rails andSeaside Sparrows (Ammodramus maritimus),but negatively correlated for Marsh Wrens(Cistothorus palustris) (Rush et al. 2009).

If no local data are available on optimaltidal stage for conducting marsh bird sur-veys, surveyors should try to conduct surveyson days when high or low tide does not fallwithin the morning (or evening) survey win-dow (i.e. conduct surveys when tides arecoming in or out). Record the following: 1)


time of the closest high tide (either the hightide before or after the survey - whichever iscloser in time) for each survey point, and 2)tidal amplitude (difference in water level inmeters between the highest and lowest tideon that day) on the day of the survey. Thesetidal features have been shown to influencenumbers of birds detected during marshbird surveys (Nadeau et al. 2010).

Survey Methods

These standardized survey methods formarsh birds originated from suggestionsduring two multi-agency workshops at Patux-ent Wildlife Research Center designed to aidagencies developing marsh bird monitoringprograms (Ribic et al. 1999; U.S. Fish andWildlife Service 2006), and incorporate sug-gestions from Conway and Gibbs (2001,2011) and recent methodological advancesin estimating detection probability and ob-server bias (Nichols et al. 2000; Farnsworth etal. 2002; MacKenzie et al. 2002). Becausemany marsh birds are secretive, seldom ob-served and vocalize infrequently, the Stan-dardized North American Marsh Bird Moni-toring Protocol instructs surveyors to broad-cast calls to elicit vocalizations during surveys(Gibbs and Melvin 1993; Conway et al. 2004;Conway and Gibbs 2005; Conway andNadeau 2010). However, because we want toestimate detection probability, estimate den-sity using distance estimators, analyze datawithout the biases associated with call-broad-cast (Conway and Gibbs 2001, 2011) and sur-vey (for some participants) non-focal spe-cies, surveyors will also record birds during a5-minute passive period prior to broadcast-ing marsh bird calls. Hence, surveyors willrecord all focal species (Appendix 1) detect-ed during both a 5-minute passive period pri-or to broadcasting recorded calls and duringa period in which pre-recorded vocalizationsof focal marsh birds are broadcast into themarsh.

The recorded calls should be obtainedfrom the Marsh Bird Survey Coordinator(contact info below); request digital record-ings of the focal species that breed in your ar-ea, and we will ensure that the broadcast se-

quence coincides with this protocol. Thebroadcast sequence should include exactly30 seconds of calls of each of the focal marshbird species that are expected breeders inyour area interspersed with 30 seconds of si-lence prior to the next focal species’ calls.The 30 seconds of calls consist of a series ofthe most common calls for that species inter-spersed with approximately 5 seconds of si-lence. For example, an entire survey se-quence might look like this:

5 minutes of silence (include a verbal state-ment at the end of each minute to alertsurveyors)

30 seconds of calls of first focal species con-figured thus:

Three Least Bittern coo-coo-coo calls

Six seconds of silence

Three Least Bittern coo-coo-coo calls

Six seconds of silence

Four series of Least Bittern kak calls

30 seconds of silence

30 seconds of calls of second focal speciesconfigured like this:

Two Sora whinny calls

Five seconds of silence

Three Sora per-weep calls

Five seconds of silence

Four Sora keep calls

30 seconds of silence

30 seconds of calls of third focal species, etc.

Include a verbal “stop” at end of the final30 seconds of silence so that surveyors knowwhen to stop the broadcast (and stop the sur-vey at that point).

Broadcast Equipment and Placement

The broadcast player should be placedupright on the ground (or on the bow of theboat), and sound pressure should be 80-90dB at 1 m in front of the speaker. Use asound-level meter to adjust volume of thebroadcast player at the beginning of eachday. If sound quality distorts when volume on

326 WATERBIRDS

your broadcast equipment reaches 80-90 dB,you should obtain higher-quality broadcastequipment. If the ground is wet, place thespeaker on an object as close to the groundas possible. Surveyors should stand 2 m toone side of the speaker while listening for vo-cal responses (standing too close to thespeaker can reduce the surveyors’ ability tohear calling birds). Surveyors should pointthe speaker toward the center of the marshand should not rotate the speaker during thecall-broadcast survey. Speakers should bepointed in the same direction for all repli-cate surveys. At points where the direction inwhich to point the speakers is not obvious(i.e. on a road or in a canal bisecting twomarshes), surveyors should record the direc-tion of the speakers at each point on a mapand on their datasheets and refer to this in-formation on all replicate surveys. Visit theprogram website to see a list of suitableequipment for broadcasting calls: http://www.cals.arizona.edu/research/azfwru/Na-tional MarshBird/).

Species to Include in the Survey Effort

Surveyors must make three decisions re-garding the species to include in their surveyeffort: 1) which species will be recorded ontheir datasheet, 2) of those species recorded,which species will be recorded during theone-minute segments (i.e. each individualbird of these species will be recorded on aseparate row on the datasheet), and 3) ofthose species recorded, which species’ callswill be included in the call-broadcast se-quence (Appendix 3). Staff from the U.S.Fish and Wildlife Service’s National WildlifeRefuge System have provided guidance onmaking these decisions for the >500 refugesin the U.S. The program website providesthis guidance along with a map overlayingthe breeding range of each focal species byUSFWS Region (http://www.cals.arizo-na.edu/research/azfwru/NationalMarsh-Bird/, then click on Focal Species and thenbreeding distribution). Surveyors should exam-ine this map to help determine which focalspecies likely breed in their area and use thisinformation to determine the species to in-

clude in the broadcast sequence. For generalinventory to document status and distribu-tion, one would include all possible focalspecies in their area.

Species to Include in the Call-broadcast Se-quence

In general, surveyors should include intheir survey all of the following species thatare thought to breed in the marshes in theirarea: King Rail, Clapper Rail, Virginia Rail,Sora, Black Rail, Yellow Rail, American Bit-tern, Least Bittern, Pied-billed Grebe,Limpkin, American Coot, Purple Gallinuleand Common Moorhen. The number ofspecies included on the call-broadcast por-tion of the survey increases the duration ofthe survey by 1 min per species at eachpoint. So, with eight species, you will spend13 minutes (including the initial 5 min pas-sive listening period) at each point. If a sur-veyor is within the breeding range of theAmerican Coot, Common Moorhen orPied-billed Grebe, broadcasting calls of anyof these species is considered optional butstrongly recommended (Appendix 3).However, all surveyors should still record alldetections of these species (see Appendix3), even if they do not include one (or allthree) of these “focal” species in theirbroadcast sequence. Participants conduct-ing surveys on National Wildlife Refugesshould see the guidance on the programwebsite on which focal species to include intheir broadcast sequence for their refuge.Use that list only as guidance and contactthe Program Coordinator (see end of docu-ment) if you can recommend corrections tothe list of suggested species for a particularrefuge. The guidance on the website isbased on maps of the breeding ranges of thefocal species, but these maps are not alwaysaccurate in identifying the species thatbreed locally at a particular refuge. If some-one has good local knowledge of the speciesof secretive marsh birds that breed on a par-ticular refuge, they should contact the Pro-gram Coordinator to have the species listupdated or verified (even if no changes arenecessary). As that occurs, we will update


the website to indicate which refuges havehad their list of species verified. Keepingthe list of recommended species on thewebsite up-to-date is important, so pleaseverify the list of focal species for your sam-pling location by asking people who are fa-miliar with the marsh birds in your area andsend any suggested modifications to theProgram Coordinator.

The broadcast sequence includes callsof the focal marsh bird species that are ex-pected breeders in that area. The calls arebroadcast on a portable CD or MP3 playerwith amplified speakers attached. Themarsh birds included in the call-broadcastsequence will vary among survey areas, butwill always be consistent within a particularsurvey route across years. Recommendedspecies to include in the call-broadcast se-quence for a particular survey route (i.e. ata particular refuge or management area)are listed in Appendix 3. The goal is to in-clude all of the focal species believed to bepotential local breeders (species for whichyou might reasonably expect to get respons-es during the breeding season). Order ofcalls start with the least intrusive speciesfirst and follow this order: Black Rail, LeastBittern, Yellow Rail, Sora, Virginia Rail,King Rail, Clapper Rail, American Bittern,Common Moorhen, Purple Gallinule,American Coot, Pied-billed Grebe,Limpkin. The order of species on thebroadcast sequence was based on recom-mendations by Ribic et al. (1999). The callsused for broadcast include the primary ad-vertising call of each species (e.g. ‘whinny’for Sora, ‘grunt’ for Virginia Rail, ‘clatter’ forClapper Rail, ‘click-click-click-click-click’ forYellow Rail, ‘coo-coo-coo’ for Least Bittern,‘pump-er-lunk’ for American Bittern). Othercalls associated with reproduction are alsoincluded for many of the species. Includingall the common calls associated with repro-duction of each species on the broadcast se-quence will likely increase detection proba-bility during different times of the breedingseason and can help surveyors learn the lesscommon calls of each of the focal species. Alist of common calls for each focal species isprovided in Appendix 4.

Estimating Distance to Each Focal Bird

Surveyors should estimate the distancefrom the survey point to each individualbird. Recording distance to each individualbird will allow analysts to use distance sam-pling techniques to estimate density for eachspecies in each habitat type and for each sur-veyor. Surveyors need to estimate the dis-tance to each bird when the bird was first de-tected during the survey. Several authorshave suggested that secretive marsh birds of-ten move toward the broadcast source priorto vocalizing (Legare et al. 1999; Erwin et al.2002), and systematic movement toward thesurveyor violates an important assumption ofdistance sampling. More research is neededto address the magnitude of this potentialproblem for each focal species, but analystswill likely use distance estimates only frombirds detected during the initial passive por-tion of the survey (i.e. those that were detect-ed prior to being exposed to call-broadcast).Estimating density from only a subset ofbirds detected (those initially detected dur-ing the 5-min passive period in this case)does not introduce bias in distance samplingas long as the other assumptions of distancesampling are met (Buckland et al. 2001).Density indices by habitat type are useful be-cause they allow managers to extrapolatesurvey data to estimate a minimum numberof each marsh bird species on their entiremanagement area. The distance at whichmost individuals are detected varies amongthe focal species (Conway and Nadeau2006). Surveyors are encouraged to use arange finder to help them determine the dis-tance to specific landmarks surroundingeach survey point, which will help estimatethe distance to calling marsh birds. Othermethods for improving one’s ability to esti-mate distance include: 1) tying surveyors’flagging at 50m and 100m away from eachsurvey point in each cardinal direction, and2) carrying aerial photos of the marsh with50m-, 100m- and 200m-radius circles drawnaround each survey point. Estimating thedistance to some individual birds will involvea lot of uncertainty (i.e. estimating distanceto birds 5m from the surveyor is much easier

328 WATERBIRDS

than estimating distance to birds that are>100m away). Indeed, distance sampling of-ten led to overestimates of density of song-birds in eastern deciduous forest (Alldredgeet al. 2007b, 2008). However, some surveyors(those that had received prior training in dis-tance estimation to calling marsh birds) wereable to estimate distance to calling secretivemarsh birds that they could not see with min-imal bias (average difference between esti-mated and actual distance varied among thefocal species from 0 to 24 m in Arizona andCalifornia based on 29-115 trials per species;Nadeau and Conway, unpublished data).More studies are needed to estimate the biasand accuracy of surveyors’ distance estimatesto calling marsh birds and the factors that af-fect bias and accuracy. Surveyors should en-ter on the datasheet and in the databasewhich of the following distance estimationaids they used: Unaided, Distance Markers,Range Finder, Range Finder and Maps, Mapsor Aerial Photos or Distance Not Recorded.

The Data Sheet

An electronic copy of a datasheetshould be obtained from the Survey Coor-dinator or the program website (http://www.cals.arizona.edu/research/azfwru/Na-tionalMarshBird/) to ensure that all perti-nent data are recorded properly. Thesedatasheets can then be tailored by each sur-veyor to meet local needs as long as none ofthe standards in this protocol are compro-mised. The number of species columns onthe datasheet will differ across survey areas.For example, if calls of only three specieswill be broadcast, then the survey durationwill be eight minutes at each point (fiveminutes of passive listening and one minuteof call-broadcast for each of three species)and will need a datasheet with eight re-sponse columns. If calls of five species willbe broadcast, then the survey duration willbe ten minutes at each point (five minutesof passive listening and one minute of call-broadcast for each of five species) and willneed a datasheet with ten response col-umns. A hypothetical example of a complet-ed datasheet is provided in Appendix 5. Pri-

or to the beginning of the survey, write downthe day, month and year at the top of the da-ta sheet. Write out the month or use a three-letter abbreviation to avoid confusion be-tween day and month (i.e. so that 6 May isnot confused with 5 June). Also write the fullname of all persons present during the sur-vey. If more than one person was present,write down who recorded the data and allpersons that helped identify calling birds.Because detection probability can differ sub-stantially among surveyors (Kendall et al.1996; Link and Sauer 1998; Conway et al.2004; Sauer et al. 2004), analysts may wish tocontrol for observer bias when estimatingtrend (similar to the approach used by ana-lysts of BBS data; Sauer et al. 2008), so partic-ipants should record any and all surveyorswho contributed to marsh bird detections(see the multiple-observer surveys section).Write down the name of the survey route andthe name of the survey area or managementunit.

Recording Detections of Focal, Broadcast Species

When you arrive at the first survey point,write down the unique identification num-ber of the survey point and the time. Startthe survey. When an individual of a focal spe-cies is detected, write the species name in the“Species” column. You can use the four-letteracronym for the species or write the full spe-cies name. A list of standard four-letter spe-cies acronyms is provided in Appendix 1. Puta “1” in each detection column in which thatindividual is detected aurally and put an “s”in each column in which the individual is de-tected visually (including flying overhead).For example, if an individual Virginia Railvocalizes during the first one minute of pas-sive listening, put a “1” in the first column.Regardless of whether that individual callsonce or many times during the first minute,only put one “1” in the first column. If thatsame individual bird is still calling during thesecond minute of passive listening, then alsoput a “1” in the second column. If the sameindividual calls during the 30 seconds whenSora calls are being broadcast or the 30 sec-


onds of silence immediately following theSora sequence, put a “1” in the column for“SORA”. If that same individual bird callsagain during the Virginia Rail sequence, alsoput a “1” in the “VIRA” column, and so on.Hence, if an individual bird is calling con-stantly throughout the survey period, youwill have a “1” in every column for that indi-vidual. If the individual is heard and seen,put both a “1” and an “s” in the appropriatecolumn(s). If you hear a call of the same spe-cies but from a different individual (or froman individual of another species), start a newrow on the data sheet and follow the sameprotocol just described. Recording whethereach individual bird is or is not detected dur-ing each 1-min segment allows analysts to useremoval models or time-of-detection meth-ods (Farnsworth et al. 2002; Alldredge et al.2007a) to estimate detection probability (seeConway et al. 2010 for an example). Survey-ors may have difficulty determining whethera call is coming from a new individual or anindividual detected earlier at that surveypoint. Surveyors must often decide whethera call is that of a new individual or one al-ready detected without seeing the bird by us-ing their best judgment (this is a challengeon all bird surveys regardless of the protocolused). In general, be conservative and as-sume that a call is from the same bird if thecall came from the same general location(i.e. a similar direction and not too far fromthe location of the original call). The num-ber of rows filled out on the datasheet willdiffer among survey points and will corre-spond to the total number of individual focalmarsh birds detected at each point. If nomarsh birds are detected at a survey point,record the point number and starting time,and write “no birds” in the Species column. Asample datasheet is included as an exampleof what survey data might look like (Appen-dix 5). If the surveyor hears a marsh bird butis unsure of its identity, the surveyor shouldwrite “unknown” in the Species column andrecord all data for this individual as de-scribed above. Make a verbal description ofthe unknown call in the Comments column(e.g. ‘soft “kak-kak-grr” - sounds like BLRAbut harsher’). Descriptions of unknown calls

are for your own use (not entered into data-base) and will aid your future identificationof unknown calls if that call is heard repeat-edly. If time permits, the surveyor can returnto the point with another expert birder whomay be able to help identify that “unknown”bird or with sound recording equipment sothat they can send the recording of the callto the Program Coordinator for identifica-tion. Some species of marsh birds give pairedduets and surveyors can often distinguishpairs of birds during surveys. Always recordeach member of a pair on its own individualrow of the datasheet. Record “pair” in theComments columns for each of the two birdsthat are thought to be members of a matedpair.

Recording Detections of Focal, Non-broad-cast Species

Whenever possible, these species (seeAppendix 3) are recorded the same way as‘focal, broadcast species’ above, but theircalls are not broadcast during the call-broad-cast portion of the survey. If surveyors areoverwhelmed by the number of focal birdsdetected, then they record these species dif-ferently than the focal, broadcast species(see the What to do if the surveyor becomes over-whelmed with too many detections section).

Recording Detections of Non-focal Species (OPTIONAL)

We recommend that surveyors do notrecord non-focal species during surveys (alsosee Johnson et al. 2009). However, some sur-veyors will want to record all species detected(including passerines, waterfowl, raptors,etc.) or perhaps a subset of all species detect-ed (i.e. include marsh-dwelling passerines,wading birds but not other species) duringtheir marsh bird surveys. Surveyors need tobe certain that they focus their attention onthe focal marsh birds (especially in areaswhere densities of secretive marsh birds arerelatively high). If a surveyor feels stronglythat they must record species in addition tothe focal species listed in Appendix 3, theshared database can accommodate these data

330 WATERBIRDS

but data on non-focal species should be re-corded differently. At each point, record thetotal number of each non-focal species de-tected within each of three distance catego-ries (≤50m, 51-100m, and >100m). Individu-al birds of non-focal species do not receivetheir own line on the data sheet and survey-ors do not need to record detections of non-focal birds using the 1-min segments (Con-way and Droege 2006).

The non-focal species included by a sur-veyor will depend on the marsh birds of in-terest at that refuge, management area orphysiographic region. For example, survey-ors may want to include non-focal specieswhich are thought to be declining or whichare not sampled well by other survey efforts.However, analysts will need to know whichadditional species were being recorded sothat these data are meaningful (i.e. if no YH-BLs are recorded at a point, analysts need toknow whether a surveyor detected zero YH-BLs or merely did not record YHBLs on theirsurvey). Therefore, each surveyor must en-ter in the database their list of “non-focal”species that they were recording during theirsurvey. Please take into consideration thatthe number of “non-focal” species includedin your survey effort may reduce your abilityto record all the relevant data for the 26 focalspecies (Appendix 1) that are the focus ofthe Standardized North American MarshBird Monitoring Protocol. Moreover, manyof the non-focal species may be adequatelysampled already by the North AmericanBreeding Bird Survey. Indeed, Johnson et al.(2009) cautioned against surveyors record-ing non-focal species for fear that the survey-or would miss focal species while paying at-tention to non-focal species.

Record Types of Calls Given

Knowing seasonal patterns of differentcall types in a local area provides useful infor-mation. For example, the frequency of differ-ent calls given (e.g. single clatter, paired clatter,kek or kek-burr for a Clapper Rail) variesthroughout the season (Conway et al. 2004;Conway, unpubl. data). Frequency of differ-ent calls given may also vary across regions.

Different call types have different functions(see Appendix 4) and can indicate pairingstatus and stages of the nesting cycle in a localarea (allowing refinement of local survey win-dows). Moreover, detection probability andobserver bias differ with different call types(e.g. Least Bittern ‘kak’ and the first part of aVirginia Rail ‘tick-it’ can be confused withClapper Rail ‘kek’ calls) and accuracy of dis-tance estimation may vary with call type (Con-way and Nadeau 2006; Conway, unpubl. da-ta). Hence, estimates of population trendsbased on data from only those call types thathave low observer bias might increase powerto detect true population trends due to reduc-tions in temporal variation in counts. Furtherresearch is needed to determine the extentthat accuracy of trend estimates can be im-proved (if at all) by including only those birdsthat gave certain types of calls, but there areadditional benefits of having observers recordcall types on surveys (Conway and Gibbs 2001,2011). For these reasons, surveyors shouldrecord all types of calls given for each focalmarsh bird detected in the Calls column onthe datasheet (see sample datasheet; Appen-dix 5). Refer to the program website to listento examples of each common call type: http://www.cals.arizona.edu/research/azfwru/Na-tionalMarshBird/.

Birds Detected at a Prior Survey Point or be-tween Points

If surveyors detect a new bird immediate-ly after the survey period at a particular point(or while walking between points) they canrecord these birds on a separate row andrecord “yes” in the Outside Survey column.Recording birds detected outside of the ac-tual survey period may be useful informationfor some of the focal species that are partic-ularly rare at a local site (species for whichdetections during the actual surveys arerare). If a surveyor detects a focal bird dur-ing a survey and the surveyor believes thatthe call was given by the same individual birdwhich was detected and recorded at a previ-ous survey point, the surveyor should recordall the relevant data for that bird and thenenter a “Yes” in the Detected at a Previous Point


column on the datasheet. When in doubt, beconservative as to whether an individual birddetected at the current point was the sameindividual recorded at a previous point (i.e.record “Yes” when in doubt).

Recording birds that were detected out-side of the standardized survey times (i.e.outside of the 10-min survey at a point) canbe useful because these birds are secretiveand rarely vocalize. For inventory purposes,surveyors may not want to ignore these de-tections, especially if, for example, they rep-resent the only Black Rail detected all day orall year. However, a problem arises if one ofthese birds detected outside of the standard-ized survey period is then detected at a sub-sequent point during the standardized sur-vey period. For example, if: 1) the surveyordetects a Black Rail after the 10-min surveyperiod at point #3 and records that bird onits own row on the datasheet (and writes“No” in the Detected at a Previous Point columnand “yes” in the Outside Survey column), and2) the surveyor then detects the same BlackRail during the 10-min survey period atpoint #4. Recording “Yes” in the Detected at aPrevious Point column for the entry at point#4 creates a problem because: For manyanalyses (including estimates of populationtrend) analysts may want to ignore all entriesthat have a “Yes” in the Detected at a PreviousPoint column and all entries that have a “Yes”in the Outside Survey column so that all indi-vidual birds are counted only once. In thescenario above, the Black Rail would havebeen ignored altogether from trend analy-ses. Hence, for the situation describedabove, the surveyor should write “No” in theDetected at a Previous Point column for the en-try at point #4, and then go back and changethe “No” to “Yes” in the Detected at a PreviousPoint column for the initial entry for this birdat point #3 (when the bird was detected afterthe 10-min survey period).

Recording Whether Focal Birds are Within the “Target Area”

A common goal of marsh bird surveys isto document the effects of management ac-tions on marsh birds, but adjacent areas may

have received different management ac-tions. Two adjacent areas with different man-agement histories presents a problem ifsome birds detected at a survey point arewithin one area but others are within anoth-er area (with a different management histo-ry). For clarity, some participants who usethis protocol will have certain survey pointsthat were located specifically to count birdswithin a certain “target area”. For example,four points along a survey route were locatedwith the intent to count marsh birds within“impoundment A” and five points along thatsame survey route were located with the in-tent to count marsh birds within “impound-ment B”. However, surveyors at these ninepoints detect birds both within these im-poundments and also in adjacent areas out-side these impoundments. Participants oftenwould like to count all birds detected at eachpoint, but also delineate which ones werewithin these impoundments. Hence, survey-ors should record, to the best of their ability,whether each bird detected was or was not intheir “target area”. In the example above,the “target area” is impoundment A forpoints 1-4 and the “target area” is impound-ment B for points 5-9. If some (or all) of yourpoints are associated with a “target area”, thename of that target area should be identifiedin the database for each point. Some partici-pants may not have any “target area” associ-ated with any of their survey points, and cantherefore leave the “Target Area” columnblank on the datasheet (and in the data-base). For example, you may not have any ex-plicit “target areas” associated with any ofyour survey points if you are conducting sur-veys primarily to estimate population trend,determine status and distribution, or toidentify habitat relationships. If the marshesyou survey undergo different managementactions, then identifying a distinct “target ar-ea” associated with each survey point may beuseful, whereby surveyors can record wheth-er each bird they detect at each point was orwas not within that “target area”. The “targetarea” may be different at each survey pointalong a survey route. If a participant has dif-ferent management units or specific marsh-es that some of their survey points are in-

332 WATERBIRDS

tended to monitor (or if the participant seesvalue in differentiating among units ormarshes), they should add a column to theirdatasheet titled “In target area” and record a“Y” or an “N” for each focal bird detected(see Appendix 5).

If the Surveyor Becomes Overwhelmed with too Many Detections

Because many of the focal species occurat relatively low densities through much oftheir range, many surveyors will detect fewor no individual birds at any given surveypoint. However, some survey points within asurvey area will have so many marsh birdscalling that surveyors will have difficulty re-cording data during each 1-min segment inwhich each individual focal bird is detected.For example, a surveyor may see or hear >20American Coots at one survey point. Whenmany birds are calling simultaneously, sur-veyors can have difficulty 1) deciding wheth-er they are hearing new individuals or previ-ously-detected ones, 2) writing new individ-uals on a new line of the datasheet, and 3)finding the correct line on which they hadnoted previously-detected birds. In these sit-uations, here are a few comments, observa-tions and suggested remedies. First, individ-ual surveyors do improve, with practice, atrecording the required information andmaking the necessary decisions even withrelatively high numbers of calling birds at apoint. However, everyone has a threshold atwhich the numbers of calling marsh birdsget too high at a particular point. Exceedingone’s ‘tolerance threshold’ occurs more fre-quently when a surveyor has many species intheir call-broadcast sequence (and hencemany detection columns on theirdatasheet). Below is a list of solutions forwhen a surveyor becomes overwhelmed, indecreasing order of preference (try thosenearer to the top of the list before resortingto those nearer to the bottom). A surveyoroften does not know until after the surveyhas started at a particular point that (s)he isbecoming overwhelmed and is not effective-ly assigning the correct calls to the correctcolumns (individuals).

1) Include a circle on each row of thedatasheet and make a ‘tick’ on each circleidentifying the general direction of thatindividual (the ticks on a circle, alongwith the distance estimate, will help youdifferentiate one individual from otherindividuals of that species as more are de-tected at that point - see the column titled“Direction” in Appendix 5).

2) If the problem is common on a particu-lar survey route, reduce the number ofspecies in your call-broadcast sequence.For example, eliminate American Coots,Common Moorhens and Pied-billedGrebes from your call-broadcast se-quence so that you have fewer species onyour call-broadcast sequence. In otherwords, still record data for all individualsof all focal marsh bird species in thesame way, but just reduce the number ofcolumns on the datasheet (and length ofthe call-broadcast sequence).

3) For those focal species that are of lowermanagement or conservation interest inyour survey area (e.g. American Coots,Common Moorhens, Pied-billedGrebes), simply write down an estimateof the total number of individuals de-tected within each of three distance cat-egories (<50m, 51-100m, >100m) forthat particular species at that point (e.g.write “AMCO: 0; 12; 23” on one line ofthe data sheet - see example on sampledata sheet attached; Appendix 5). Usethe 1-min segments only for the focalspecies of higher management concern(e.g. Black Rails, Yellow Rails, King Rails,Clapper Rails, bitterns). Surveyors mustrecord on the datasheet (and in the da-tabase) the points at which they wereoverwhelmed and could not record datafor individual birds on separate rows ofthe datasheet (for focal species).

4) At points at which the surveyor becomesoverwhelmed, the surveyor should en-sure that all individuals detected are re-corded on the datasheet even if thatmeans failing to record distance esti-mates to each focal bird and failing to in-clude each bird on a separate line (i.e. at


a minimum, surveyors always need toprovide an estimate of the total numberof individuals detected at each point foreach of the focal species).

Distinguishing King Rails from Clapper Rails

King Rails breed in freshwater marshesand Clapper Rails breed in saltwater marshes(except the Yuma Clapper Rail that breeds infreshwater marshes in Arizona and Califor-nia; Conway et al. 1993). Both species havesimilar calls. In marshes near coastal areas,surveyors may not be able to determinewhether birds heard calling are King Rails orClapper Rails. In those situations, surveyorsshould record these individuals as KCRA(King-Clapper Rails).

Recording Ambient Noise Level

Surveyors should record the level ofbackground noise during the survey at eachsurvey point. Background noise can be usedas a covariate in future analyses because levelof background noise varies spatially and tem-porally and influences detection probabilityduring bird surveys (Pacifici et al. 2008). Cat-egorize the background noise at each pointon a scale from 0 to 4 (or 9):

0 = no background noise during virtually all ofthe survey,

1 = faint background noise during at least halfof the survey,

2 = moderate background noise (probably can-not hear some birds beyond 100m during>30 seconds of the survey),

3 = loud background noise (probably cannothear some birds beyond 50m during >30 sec-onds of the survey),

4 = intense background noise (probably cannothear some birds beyond 25m during >30 sec-onds of the survey).

9 = not recorded.

Weather Restrictions

Weather can affect detection probabilityof marsh birds (Conway and Gibbs 2001,

2011). Surveys should only be conductedwhen wind speed is <20 km/hr (5.5 m/sec;12 mph), and not during periods of sus-tained rain or heavy fog. Even winds <20km/hr affect the detection probability ofmarsh birds (C. Conway, pers. obs.). Survey-ors should postpone surveys if they believewinds are affecting calling probability ofmarsh birds. Recommendations for conduct-ing surveys in very windy locations include:

1) Determining what time(s) of day has theleast wind in your area. The daily surveywindows in the protocol are recommen-dations; survey times should be modi-fied under conditions where wind regu-larly affects vocalization frequency or ob-server detection rates. It is importantthat surveys be conducted during thesame daily time window each year at aparticular location, and that survey win-dows at a particular location should be atthe time of day or night which has thehighest detection probability for the fo-cal species present in your area. In somelocations, surveys conducted after sunset(or before sunrise) may have higher de-tection probability compared to themorning and evening survey windowsrecommended in the protocol becausestrong winds may be less frequent dur-ing the middle of the night. In these sit-uations, surveys should be conducted atnight.

2) Trying to be flexible with your schedule,if you can. For example, plan to conducta survey on a particular day but post-pone to the following day if wind affectscalling behavior (or observers’ ability tohear calls), and keep postponing untilyou get a day that meets the acceptableweather criteria to complete the survey.

If wind speed increases to above 20 km/hr during the survey (or sustained rain be-gins while the survey is already underway),surveyors have two options: 1) stopping thesurvey and repeating the entire survey routeanother day, or 2) returning within the nextfive days to resume the survey at the time itwas stopped to repeat the remaining points

334 WATERBIRDS

on the route. Repeating the entire route ona day with better weather conditions will like-ly reduce annual variation in detection prob-ability and increase the accuracy of trend es-timates because most of the focal speciesstop calling entirely with even moderatewind speeds (and hence detection probabil-ity drops to almost 0%).

Weather Conditions

Record ambient temperature, wind speedand sky condition at each survey point.Weather variables (wind speed, in particular)influence detection probability of the focalspecies and can change dramatically duringthe 2.5 hours typically required to complete asurvey route. Use the same wind speed codesand sky condition codes as the North Ameri-can Breeding Bird Survey. Record a BeaufortNumber (0-5) on data sheet, not mi/hr orkm/hr, for wind speed (National WeatherService 2011). Record whether you measuredthe ambient temperature in degrees Celsius(oC) or degrees Fahrenheit (oF). Record theappropriate National Weather Service skycode (0-9) on the data sheet to document theextent of cloud cover and presence of fog orprecipitation (North American BreedingBird Survey 2011).

Water Conditions Associated with each Sur-vey Point (or each Management Unit)

Water level influences abundance and dis-tribution of marsh birds (Conway et al. 1993;Eddleman et al. 1994). Water levels vary annu-ally and even daily in some marshes and thesefluctuations can explain spatial and temporalchanges in marsh bird abundance. Some Na-tional Wildlife Refuges control water levels insome of their management units and have theability to directly affect marsh birds via watermanagement. Hence, surveyors should try toplace one or more water gauges for measur-ing water level in permanent locations atpoints that have the same hydrologic regime(i.e. the same daily and annual fluctuations inwater level) as the marshes being surveyed. Ifall marshes along a survey route are subject tothe same hydrologic regime (i.e. all survey

points are in the same river system or are in asingle management unit with the same hydro-logic regime), then only one water gauge isneeded for that entire route. If a survey routehas points split between ≥2 managementunits (or ≥2 areas with different hydrologic re-gimes), then ≥2 water gauges are necessaryand surveyors should record on the data sheetthe water gauge associated with each surveypoint. Water level at each water gauge shouldbe recorded immediately before or immedi-ately after a morning or evening survey routeis completed. Each water gauge must be “re-set” (recalibrated) each year because freezingand thawing can cause gauges to move lateral-ly. Water gauges should be placed in an areawhere the water is deepest to avoid zero read-ings when there is still water in other parts ofthe marsh. These water gauges are not meantto explain differences in birds detectedamong points along a route, but rather thereadings from these water gauges will help ex-plain variation in numbers of marsh birds de-tected across years and across the three annu-al surveys of each route. Hence, these waterdepths can be used as covariates in many anal-yses of marsh bird survey data (i.e. in esti-mates of trend). Surveyors should record thetype of water gauge used for measuring waterdepth (i.e. bathymetry, monitoring wells, pie-zometer, river readings at gauge, staff gaugestuck into the wetland, etc.). Placing watergauges might not be possible on all surveyroutes, especially those that survey wetlandson private lands, but permission should besought from land owners or land manage-ment agencies to place water gauges on asmany survey routes as possible. Water depthcan vary widely from year to year in many wet-lands, and changes in water depth have tre-mendous effects on suitability of a marsh forparticular species. Hence, any efforts to quan-tify annual changes in water depth will dra-matically improve an analyst’s ability to esti-mate trends (and help explain the cause ofsome trends).

Salinity Content of Water

In coastal marshes or any marshes withvarying salinity levels, surveyors are encour-


aged to record the salinity content of the wa-ter directly in front of each point on eachsurvey. Salinity levels affect habitat suitabilityfor many species of marsh birds (i.e. KingRails are thought to occur only in marsheswith low salinity and Clapper Rails only inmarshes with high salinity) and such infor-mation is relatively easy to collect and can beused to help shed more light on the effect ofsalinity on distribution and abundance ofmarsh birds. Moreover, salinity levels incoastal marshes may change with changes insea levels as a result of climate change, sodocumenting changes in salinity levels overtime will help document the effects of sealevel rise.

Date of Last Natural Disturbance

Record the month and year of the lastflood, wild fire, hurricane, monsoon, torna-do, straightline winds or other major distur-bance that occurred in the “target area” thateach survey point is intended to survey.Record the month and year for each of thosedisturbance events that potentially affectedmarsh bird abundance or marsh bird habitatstructure. Record these dates for each surveypoint, once per year (or more often if a nat-ural disturbance occurs between two repli-cate surveys during the same year).

Select the choice that best describes thedisturbance that occurred at the point. Thefollowing is a list of the choices:

Date of Last Management Action

Many common wetland management ac-tivities (e.g. prescribed fire, drawdown,flooding, disking, mowing, grazing, herbi-cide application) may affect abundance of

marsh birds. For example, periodic burningof emergent marshes benefits Clapper Railsand Black Rails along the Lower ColoradoRiver (Conway et al. 2010). Hence, surveyorsshould record the month and year of the lasttime each of the above management activi-ties occurred in the 50-m radius area sur-rounding each survey point. Recording thedate of past management activities will allowanalysts to evaluate the effects of commonmanagement actions on marsh bird abun-dance at a large (continental) spatial scalewith the pooled data.

Inclusion of an Initial Settling Period (NOT RECOMMENDED)

When surveyors are using a motorizedboat or airboat to travel between surveypoints, the noise generated by the boat maycause birds to stop calling. In these situa-tions, surveyors may choose to include a “set-tling” period of a fixed length of time (e.g.one minute) prior to starting the 5-min pas-sive count at each point. Otherwise, we rec-ommend that no settling period be includ-ed. If a surveyor includes an initial settlingperiod prior to each survey, the surveyorshould keep that settling period constant forall future surveys at those points. If included,make the settling period a part of a writtensurvey protocol for future surveyors and partof the datasheets for that site so that individ-uals wishing to repeat the effort in futureyears will know that a settling period was in-cluded.

Multiple-observer Surveys (OPTIONAL)

Estimating detection probability associat-ed with a particular survey protocol is helpfulwhen attempting to interpret count data pro-duced from a monitoring program. The ex-tent to which trends in count data representthe underlying trend in true abundance de-pends on the variation in detection probabil-ity and magnitude of observer bias. One wayto estimate observer bias associated with oursurvey effort is via a modification of the dou-ble-observer method (Nichols et al. 2000).Two or more trained surveyors record data

building construction drained (drained wetland)

road construction clear-cut

fire selective harvest

chemical treatment hurricane

trail construction destructive use (non-harvest)

ice damage wind event/blow down

insect damage other

336 WATERBIRDS

independently at a series of survey points (forexamples, see Conway et al. 2004; Nadeau etal. 2008). Having ≥2 observers record data in-dependently at a point is an improvementover the original method described byNichols et al. (2000) because analysts can esti-mate detection probability for both observers(and this approach does not assume that thesecondary observer is always correct if dis-crepancies occur). Nichols et al. (2000) didnot recommend this approach only becausethey were concerned that two observersstanding next to each other could not recordterrestrial birds (where many detections arevisual and surveyors often use binoculars todetect and identify birds) independently.However, independent observations are pos-sible during marsh bird surveys (C. Conway,pers. obs.) where the large majority of detec-tions are aural (Conway et al. 2004; Nadeau etal. 2008). Hence, whenever possible, surveysshould be conducted by two or more survey-ors simultaneously. Each surveyor should fillout a separate datasheet and should recordtheir data separately without discussing datawith the other surveyor. Surveyors should notpoint out a call or a bird to the other survey-ors during the survey period. Each surveyorshould stand 1-2 meters away from each oth-er and should keep their pen on their datasheet at all times so that one surveyor is notcued by the sudden writing activity of anoth-er surveyor. Once the survey for that morn-ing or evening is completed, the surveyorscan look over each other’s data and discussdiscrepancies, but the data should not be al-tered; obvious mistakes should be noted inthe Comments column but not changed (thedifferences between the surveyors in numberof birds detected at each point is what allowsanalysts to estimate surveyor bias so these dif-ferences should not be altered). The Pro-gram Coordinator can provide a dataformfor recording which birds were detected byboth surveyors and which were only detectedby one of the surveyors. Multiple-observersurveys will obviously not be possible at alltimes and at all locations, but try to use mul-tiple-observer surveys whenever possible sothat analysts can obtain sufficient data to esti-mate observer bias.

Hearing Tests (OPTIONAL)

Surveyors are strongly encouraged tohave a hearing test (audiogram) at a quali-fied hearing or medical clinic before, dur-ing, or immediately after the survey seasoneach year. We encourage surveyors or poten-tial surveyors to discuss the results of theirhearing with their doctor and with their su-pervisor (or the Program Coordinator) todetermine whether the quality of the datathey collect may be compromised. Remem-ber, ~90% of marsh bird detections are aural,and many calls are very faint or are given bybirds a long way off. For example, 43% ofClapper Rails detected during surveys alongthe Lower Colorado River were detected atdistances >100m from the surveyor. Hearingability could be included as a covariate andmight help control for observer bias in trendanalyses because many of the focal species of-ten give very faint calls that are difficult forpeople with hearing impairments to detect(C. Conway, pers. observ.).

Personnel and Training

All surveyors should have the ability toidentify all common calls of focal and non-focal marsh bird species in their local area.Regularly listening to the recorded callsused for surveys can help you learn calls,but surveyors should also practice call iden-tification at marshes (outside the intendedsurvey area if necessary) where the focalspecies are frequently heard calling. Annu-al training workshops occur, so contact theProgram Coordinator for information onupcoming training workshops. All surveyorsshould also be trained to estimate distanceto calling marsh birds, and to identify thecommon species of wetland plants withinthe survey area. Methods for training sur-veyors to accurately estimate distance in-clude: 1) place a broadcast device withspeakers in the marsh at a known distanceand have surveyors estimate distance, 2)identify a piece of vegetation in the marshwhere the bird is thought to be calling fromand use a range-finder to determine dis-tance to that plant, 3) have a surveyor esti-


mate the distance to a bird that is callingwith regularity and is near a road or marshedge, then have a second surveyor walkalong the road/edge until they are adja-cent from that calling bird, and then mea-sure this distance (by pacing or use of aGPS) and determine how accurate the sur-veyor was at estimating distance. Multiple-observer surveys (see above) are very usefulhere - after the survey is complete have thetwo surveyors discuss what they heard andtheir distance estimates to each bird. Peri-odic multiple-observer surveys not onlyprovide estimates of detection probability(see above) but also provide insights re-garding whether a surveyor might be con-sistently underestimating or overestimat-ing distance to calling birds (this informa-tion does not tell you which observer is cor-rect, but indicates that at least oneobserver is estimating with bias and henceboth should receive more training). First-time surveyors can “tag along” on surveysconducted by more experienced surveyorsin their region prior to starting their ownsurveys. They should do at least one “trialrun” before their first data collection win-dow begins because errors are less likely ifsurveyors take time to get used to the datasheet so that they can record the appropri-ate data correctly. Another training tool isthe Western Great Lakes Birder Certifica-tion Program (http://www.uwgb.edu/birds/certification/index-1.htm).

Supplies Needed for Surveys

• surveyor flagging (to mark survey points)

• GPS receiver

• clipboard, datasheets, pencils

• CD or MP3 file of focal species’ calls (ob-tained from the Program Coordinator - see contact info below)

• broadcast device

• amplified speakers

• batteries for the broadcast device and the amplified speakers

• sound level meter with ±5 dB precision

• thermometer

• water gauge(s)

• salinity meter (e.g. Portable Salinity Re-fractometer from Forestry Suppliers, $90)

Batteries should be changed or re-charged frequently (before sound quality de-clines or devices die during a survey). Sur-veyors should routinely ask themselves if thequality of the broadcast sound is high. Ob-tain a new broadcast device or a new set ofcalls if quality seems to decline. Surveyorsshould always carry replacement batteries onall surveys. A spare broadcast device shouldbe kept closeby in case the primary unit failsto operate.

Data Entry

Surveyors are urged to enter their data tothe pooled database online at: http://www.pwrc.usgs.gov/point/mb/. Each newsurvey site needs to register using this portalcurrently maintained by the U.S. GeologicalSurvey’s Patuxent Wildlife Research Centerin order to enter and manage data. The da-tabase was specifically built to accommodateprocedures of the Standardized NorthAmerican Marsh Bird Monitoring Protocoland was designed to facilitate data manage-ment. Surveyors will be able to proof theirdata on the website after it has been enteredand will be able to obtain an electronic copyof their data (in MS Access) immediately af-ter entry. The database enables efficient en-try and storage of marsh bird survey data andensures that data are entered in a consistentway that minimizes errors.

Organizational Information

Visit the National Marsh Bird MonitoringWebsite (http://www.cals.arizona.edu/re-search/azfwru/NationalMarshBird/) for ad-ditional information on these protocols, ob-taining datasheets, obtaining digital record-ings of calls and marsh bird monitoring ingeneral. E-mail the name, address, phonenumber and e-mail address of all surveyorsto the address below. An e-mail distributionlist will be used to disseminate informationregarding training workshops and other in-

338 WATERBIRDS

formation to each surveyor. Several USFWSNational Wildlife Refuges and several partic-ipants from other organizations began usingthese marsh bird survey methods in 1999(Conway and Nadeau 2006). Over 300 par-ticipants have used the Standardized NorthAmerican Marsh Bird Monitoring Protocoland submitted data to a pooled databasewhich includes data from over 80,000 marshbird surveys (Conway and Nadeau 2006). Forassistance obtaining digital calls of the focalspecies, additional information or questionsregarding standardized marsh bird surveymethods please contact:

Dr. Courtney J. ConwayUSGS Idaho Cooperative

Fish & Wildlife Research UnitP.O. Box 441141University of IdahoMoscow, ID, 83844, USAph: 208-885-6336FAX: 208-885-9080E-mail: [email protected]

ACKNOWLEDGMENTS

The Standardized North American Marsh Bird Mon-itoring Protocol underwent review by the Marsh BirdUser Acceptance Team of the U.S. Fish and Wildlife Ser-vice. Its members (S. Lor, J. Casey, H. Laskowski, T.Doyle, B. Peterjohn, K. Blair, D. Granfors, J. Holler, A.Hanley, M. Wimer, S. Kodrow and P. Hess) providedcomments to improve the protocol. D. Johnson and M.Woodrey provided suggestions for improvement. Anyuse of trade names is for descriptive purposes only anddoes not imply endorsement by the U.S. Government.

LITERATURE CITED

Alldredge, M. W., K. Pacifici and T. R. Simons. 2008. Anovel field evaluation of the effectiveness of distanceand independent observer sampling to estimate au-ral avian detection probabilities. Journal of AppliedEcology 45: 1349-1356.

Alldredge, M. W., K. H. Pollock, T. R. Simons, J. A. Col-lazo and S. A. Shriner. 2007a. Time-of-detectionmethod for estimating abundance from point countsurveys. Auk 124: 653-664.

Alldredge, M. W., T. R. Simons and K. H. Pollock.2007b. A field evaluation of distance measurementerror in auditory avian point count surveys. Journalof Wildlife Management 71: 2759-2766.

Bogner, H. E. and G. A. Baldassarre. 2002. The effective-ness of call-response surveys for detecting least bit-terns. Journal of Wildlife Management 66: 976-984.

Buckland, S. T., D. R. Anderson, K. P. Burnham, J. L.Laake, D. L. Borchers and L. Thomas. 2001. Intro-

duction to distance sampling: estimating abundanceof biological populations. Oxford Press, New York,New York.

Burnham, K. P. 1981. Summarizing remarks: Environmen-tal influences. Studies in Avian Biology 6: 324-325.

Bystrak, D. 1981. The North American breeding birdsurvey. Studies in Avian Biology 6: 34-41.

Conway, C. J. 1995. Virginia Rail (Rallus limicola). In Thebirds of North America. No. 173. (A. Poole, P.Stettenheim and F. Gill, Eds.). The Academy of Nat-ural Sciences, Philadelphia, Pennsylvania.

Conway, C. J. and S. Droege. 2006. A Unified Strategyfor Monitoring Changes in Abundance of Birds As-sociated with North American Tidal Marshes. Stud-ies in Avian Biology 32: 382-397.

Conway, C. J., W. R. Eddleman and S. H. Anderson.1994. Nesting success and survival of Virginia Railsand Soras. Wilson Bulletin 106: 466-473.

Conway, C. J., W. R. Eddleman, S. H. Anderson and L.R. Hanebury. 1993. Seasonal changes in Yuma Clap-per Rail vocalization rate and habitat use. Journal ofWildlife Management 57: 282-290.

Conway, C. J. and J. P. Gibbs. 2001. Factors influencingdetection probabilities and the benefits of call-broadcast surveys for monitoring marsh birds. FinalReport, USGS Patuxent Wildlife Research Center,Laurel, Maryland. http://ag.arizona.edu/research/azfwru/cjc/, accessed 22 March 2011.

Conway, C. J. and J. P. Gibbs. 2005. Effectiveness of call-broadcast surveys for monitoring marsh birds. Auk122: 26-35.

Conway, C. J. and J. P. Gibbs. 2011. Summary of the in-trinsic and extrinsic factors affecting detection prob-ability of marsh birds. Wetlands 31: 403-411.

Conway, C. J. and C. Nadeau. 2006. Development andfield-testing of survey methods for a continentalmarsh bird monitoring program in North America.Wildlife Research Report # 2005-11. USGS ArizonaCooperative Fish and Wildlife Research Unit, Tuc-son, Arizona. http://ag.arizona.edu/research/azfw-ru/cjc/, accessed 22 March 2011.

Conway, C. J. and C. P. Nadeau. 2010. The effects of con-specific and heterospecific call-broadcast on detec-tion probability of marsh birds in North America.Wetlands 30: 358-368.

Conway, C. J., C. P. Nadeau and L. Piest. 2010. Fire helpsrestore natural disturbance regime to benefit rareand endangered marsh birds endemic to ColoradoRiver. Ecological Applications 20: 2024-2035.

Conway, C. J. and C. Sulzman. 2007. Status and habitatuse of the California black rail in the southwesternU.S.A. Wetlands 27: 987-998.

Conway, C. J., C. Sulzman and B. A. Raulston. 2004. Factorsaffecting detection probability of California BlackRails. Journal of Wildlife Management 68: 360-370.

COSEWIC. 2002. Canadian Species at Risk, May 2002.Committee on the Status of Endangered Wildlife inCanada. Canadian Wildlife Service, Ottawa, Ontar-io, Canada.

Dahl, T. E. 2006. Status and trends of wetlands in theconterminous United States 1998 to 2004. U.S. De-partment of the Interior, Fish and Wildlife Service,Washington, D.C.

Diario Oficial de la Federacion. 2002. Norma OficialMexicana NOM-059-ECOL-2001, Proteccion ambi-ental-Especies nativas de Mexico de flora y fauna sil-vestres-Categoriosde riesgo y especificaciones parasu inclusion, exclusion o cambio-Lista de especies en


riesgo. Secretaria de Medio Ambiente y RecursosNaturales, 6 Marzo 2002.

Eddleman, W. R., R. E. Flores and M. Legare. 1994.Black Rail (Laterallus jamaicensis). In The Birds ofNorth America Online. No. 123. (A. Poole, Ed.).Cornell Lab of Ornithology, Ithaca, New York. http://bna.birds.cornell.edu/bna/species/123doi:10.2173/bna.123, accessed 22 March 2011.

Eddleman, W. R., F. L. Knopf, B. Meanley, F. A. Reid andR. Zembal. 1988. Conservation of North Americanrallids. Wilson Bulletin 100: 458-475.

Erwin, R. M., C. J. Conway and S. W. Hadden. 2002. Spe-cies occurrence of marsh birds at Cape Code Nation-al Seashore, Massachusetts. Northeastern Naturalist9: 1-12.

Farnsworth, G. L., K. H. Pollock, J. D. Nichols, T. R. Si-mons, J. E. Hines and J. R. Sauer. 2002. A removalmodel for estimating detection probabilities frompoint-count surveys. Auk 119: 414-425.

Gibbs, J. P. and S. M. Melvin. 1993. Call-response surveysfor monitoring breeding waterbirds. Journal ofWildlife Management 57: 27-34.

Gibbs, J. P., S. Melvin and F. A. Reid. 1992. American Bit-tern (Botaurus lentiginosus). In The Birds of NorthAmerica. No. 18. (A. Poole, P. Stettenheim and F.Gill, Eds.). The Academy of Natural Sciences, Phila-delphia, Pennsylvania.

Johnson, D. H. 2008. In defense of indices: The case ofbird surveys. Journal of Wildlife Management 72:857-868.

Johnson, D. H., J. P. Gibbs, M. Herzog, S. Lor, N. D. Ni-emuth, C. A. Ribic, M. Seamans, T. L. Shaffer, W. G.Shriver, S. V. Stehman and W. L. Thompson. 2009. Asampling design framework for monitoring secretivemarshbirds. Waterbirds 32: 203-362.

Kendall, W. L., B. G. Peterjohn and J. R. Sauer. 1996.First-time observer effects in the North AmericanBreeding Bird Survey. Auk 113: 823-829.

Klaas, E. E., H. M. Ohlendorf and E. Cromartie. 1980.Organochlorine residues and shell thicknesses ineggs of the Clapper Rail, Common Gallinule, PurpleGallinule, and Limpkin (Class Aves), eastern andsouthern United States, 1972-74. Pesticides Monitor-ing Journal 14: 90-94.

Lawler, J. J. and R. J. O’Connor. 2004. How well do con-sistently monitored breeding bird survey routes rep-resent the environments of the conterminousUnited States? Condor 106: 801-814.

Legare, M. L., W. R. Eddleman, P. A. Buckley and C.Kelly. 1999. The effectiveness of tape playback in es-timating Black Rail density. Journal of Wildlife Man-agement 63: 116-125.

Lewis, C. and D. G. Casagrande. 1997. Using avian com-munities to evaluate salt marsh restoration. Pages204-236 in Restoration of an Urban Salt Marsh (D.G. Casagrande, Ed.). Yale School of Forestry and En-vironmental Studies Bulletin Series 100, New Haven,Connecticut.

Link, W. A. and J. R. Sauer. 1998. Estimating populationchange from count data: application to the NorthAmerican Breeding Bird Survey. Ecological Applica-tions 8: 258-268.

MacKenzie, D. I., J. D. Nichols, G. B. Lachman, S.Droege, J. A. Royle and C. A. Langtimm. 2002. Esti-mating site occupancy rates when detection proba-bilities are less than one. Ecology 83: 2248-2255.

Meanley, B. 1992. King Rail (Rallus elegans). In TheBirds of North America. No. 3. (A. Poole, P. Stetten-

heim and F. Gill, Eds.). The Academy of Natural Sci-ences, Philadelphia, Pennsylvania.

Nadeau, C. P., C. J. Conway, M. A. Conway and J. Rein-man. 2010. Variation in Clapper Rail and Least Bit-tern Detection Probability among Tidal Stages onthe Northern Coast of the Gulf of Mexico. WildlifeResearch Report #2010-01, U.S. Geological Survey,Arizona Cooperative Fish and Wildlife ResearchUnit, Tucson, Arizona. http://ag.arizona.edu/re-search/azfwru/cjc/, accessed 22 March 2011.

Nadeau, C. P., C. J. Conway, B. S. Smith and T. E. Lewis.2008. Maximizing detection probability of wetland-dependent birds during point-count surveys innorthwestern Florida. Wilson Journal of Ornitholo-gy 120: 513-518.

National Audubon Society. 2007. The 2007 AudubonWatchlist. National Audubon Society, New York. ht-tp://birds.audubon.org/2007-audubon-watchlist,accessed 22 March 2011.

National Weather Service. 2011. Beaufort Wind Scale.National Oceanic and Atmospheric Administration,Norman, Oklahoma. http://www.spc.noaa.gov/faq/tornado/beaufort.html, accessed 31 March2011.

North American Breeding Bird Survey. 2011. U.S. Geo-logical Survey, Patuxent Wildlife Research Center,Laurel, Maryland. http://www.pwrc.usgs.gov/bbs/participate/training/ appendixA.html, accessed 31March 2011.

Nichols, J. D., J. E. Hines, J. R. Sauer, F. W. Fallon, J. E. Fal-lon and P. J. Heglund. 2000. A double-observer ap-proach for estimating detection probability andabundance from avian point counts. Auk 117: 393-408.

Odom, R. R. 1975. Mercury contamination in Georgiarails. Proceedings of the Annual Conference of theSoutheastern Association of Game & Fish Commis-sions 28: 649-658.

Pacifici, K., T. R. Simons and K. H. Pollock. 2008. Effectsof vegetation and background noise on the detec-tion process in auditory avian point-count surveys.Auk 125: 600-607.

Rehm, E. M. and G. A. Baldassarre. 2007. Temporal vari-ation in detection of marsh birds during broadcastof conspecific calls. Journal of Field Ornithology 78:56-63.

Robbins, C. S., D. Bystrak and P. H. Geissler. 1986. Thebreeding bird survey: Its first fifteen years, 1965-1979. U.S. Dept. of the Interior, Fish and WildlifeService Resource Publication 157. Washington, D.C.

Ribic, C. A., S. Lewis, S. Melvin, J. Bart and B. Peterjohn.1999. Proceedings of the marsh bird monitoringworkshop. USFWS Region 3 Administrative Report,Fort Snelling, Minnesota. http://library.fws.gov/Pubs/marsh_bird.pdf, accessed 22 March 2011.

Rosenstock, S. S., D. R. Anderson, K. M. Giesen, T. Leu-kering and M. F. Carter. 2002. Landbird countingtechniques: Current practices and an alternative.Auk 119: 46-53.

Rush, S. A., E. C. Soehren, K. W. Stodola, M. S. Woodreyand R. J. Cooper. 2009. Influence of tidal height ondetections of breeding marsh birds along the north-ern Gulf of Mexico. Wilson Journal of Ornithology121: 399-405

Sauer, J. R, J. E. Hines and J. Fallon. 2008. The NorthAmerican Breeding Bird Survey, Results and Analy-sis 1966–2007. Version 5.15.2008. USGS PatuxentWildlife Research Center, Laurel, Maryland. http://www.mbr-pwrc.usgs.gov/bbs/, accessed 5 July 2010.

340 WATERBIRDS

Sauer, J. R., W. A. Link and J. A. Royle. 2004. Estimatingpopulation trends with a linear model: Technicalcomments. Condor 106: 435-440.

Spear, L. B., S. B. Terrill, C. Lenihan and P. Delevoryas.1999. Effects of temporal and environmental factorson the probability of detecting California black rails.Journal of Field Ornithology 70: 465-480.

Stedman, S. and T. E. Dahl. 2008. Status and trends ofwetlands in the coastal watersheds of the EasternUnited States 1998 to 2004. U.S. Department of theInterior, Fish and Wildlife Service. Washington, D.C.http://www.fws.gov/wetlands/_documents/gSandT/ NationalReports/StatusTrends WetlandsCoastalWatershedsEasternUS1998to2004.pdf, accessed 22March 2011.

Tate, J. 1986. The blue-list for 1986. American Birds 40:227-236.

Thompson, W. L. 2002. Towards reliable bird surveys:Accounting for individuals present but not detected.Auk 119: 18-25

Thompson, W. L., G. C. White and C. Gowan. 1998.Monitoring Vertebrate Populations. AcademicPress, San Diego, California.

Tiner, R. W., Jr. 1984. Wetlands of the United States: cur-rent status and recent trends. U. S. Fish and WildlifeService, National Wetlands Inventory, Washington,D.C.

U.S. Fish and Wildlife Service. 2008. Birds of ConservationConcern 2008. U. S. Department of Interior, Fish andWildlife Service, Division of Migratory Bird Manage-ment, Arlington, Virginia. http://www.fws.gov/migra-torybirds/, accessed 22 March 2011.

U.S. Fish and Wildlife Service. 2005. The U.S. Fish andWildlife Service’s Focal Species Strategy for Migrato-ry Birds. Division of Migratory Bird Management,Arlington, Virginia. http://www.nbii.gov/portal/server.pt?open=512&objID=760&mode=2&in_hi_userid=2&cached=true, accessed 1 April 2011.

U.S. Fish and Wildlife Service. 2006. Proceedings of the2006 Marsh Bird Monitoring Technical Workshop,March 6-8, 2006. Patuxent Wildlife Research Center,Laurel, Maryland. http://www.fws.gov/birds/water-birds/monitoring/marshmonitoring.html, accessed22 March 2011.

Zembal, R. and B. W. Massey. 1987. Seasonality of vocal-izations by Light-footed Clapper Rails. Journal ofField Ornithology 58: 41-48.


Appendix 1. AOU 4-letter species acronyms for themarsh birds that are the focus of this protocol.

BLRA Black Rail

YERA Yellow Rail

SORA Sora

VIRA Virginia Rail

KIRA King Rail

CLRA Clapper Rail

KCRA King/Clapper Rail

YBCR Yellow-breasted Crake

LEBI Least Bittern

AMBI American Bittern

LIMP Limpkin

PUGA Purple Gallinule

COMO Common Moorhen

AMCO American Coot

CARC Caribbean Coot

PBGR Pied-billed Grebe

LEGR Least Grebe

EAGR Eared Grebe

RNGR Red-necked Grebe

HOGR Horned Grebe

CLGR Clark’s Grebe

WISN Wilson’s Snipe

BLTE Black Tern

SSTS Saltmarsh Sharp-tailed Sparrow

NSTS Nelson’s Sharp-tailed Sparrow

SESP Seaside Sparrow

WILL Willet (Eastern)

Examples of Non-focal Species these are just some exam-ples -there are other wetland birds that a surveyor maywant to include; each cooperator should decide whichnon-focal species to include in their surveys in advanceand list these species on their datasheet and in the data-base so that analysts (and surveyors in future years) willknow the list of species recorded in prior years. Onecaution to remember - choosing too many non-focal spe-cies may cause surveyors to become overwhelmed withdata collection and non-focal species should not be re-corded at the expense of data on the focal species.Once a station decides to include certain non-focal spe-cies, every surveyor at that station should record themin the same manner each year so that the data for thatspecies from that station are valid.

GRHE Green Heron

GBHE Great Blue Heron

GLIB Glossy Ibis

FOTE Foster’s Tern

SEWR Sedge Wren

MAWR Marsh Wren

LCSP Le Conte’s Sparrow

SWSP Swamp Sparrow

YHBL Yellow-headed Blackbird

342 WATERBIRDS

Appendix 2. Dates of 3 annual survey windows for different areas in North America. The isoclines are based onaverage maximum temperatures in May, from PRISM at Oregon State University (for the U.S.) and EnvironmentCanada (for Canada).

14

14

14


Appendix 3. List of the focal marsh bird species and their field data requirements for conducting marsh bird mon-itoring using the North American Marsh Bird Monitoring Protocol for NWRS. These are the species for which theMarsh Bird Monitoring Program is designed to monitor well. Surveyors should always record at least total # detect-ed at each point for all these species.

Species Broadcast Required?1 Record One Individual/Line?

BroadcastBLRA YES YESYERA YES YESSORA YES YESVIRA YES YESKIRA YES YESCLRA YES YESYBCR YES YESLEBI YES YESAMBI YES YESLIMP YES YESPUGA YES YESCOMO Recommended YES, except2

AMCO Recommended YES, except2

CARC Recommended YES, except2

PBGR Recommended YES, except2

Non-broadcastWILL (Eastern) NO YES, except2

RNGR NO YES, except2

EAGR NO YES, except2

HOGR NO YES, except2

CLGR NO YES, except2

LEGR NO YES, except2

WISN NO YES, except2

SSTS NO YES, except2

NSTS NO YES, except2

SESP NO YES, except2

BLTE NO YES, except2

1BROADCAST REQUIRED: species for which surveyors must broadcast calls if they are conducting surveys within the breed-ing range of that species. Recommended= use of broadcast is optional (BUT strongly encouraged) for these species even if surveyoris within the breeding range of that species.

2Record each individual on one row of the data form except at points where the surveyor is overwhelmed because too many focalbirds are being detected at that point (see page 332 of the protocol).

344 WATERBIRDSA

ppen

dix

4. L

ist o

f th

e m

ost c

omm

on c

alls

for

the

foca

l spe

cies

of

mar

sh b

irds

.

Spec

ies

Stan

dard

ized

Cal

l Nam

eSi

bley

Nam

e(s)

BN

A N

ame(

s)1

Poss

ible

Fun

ctio

nSa

mpl

e on

B

NA

1 web

site

AM

BI

pum

p-er

-lun

k bl

oon

k-ad

oon

kpu

mp-

er lu

nk

and

dun

k-a-

doo

mat

e at

trac

tion

, ter

rito

rial

sig

nal

AM

BI

chu-

peep

ch

u-pe

epch

u-pe

ep

duri

ng

copu

lati

on c

erem

ony

AM

BI

kok

kok-

kok-

kok

kok-

kok-

kok

or h

ain

k w

hen

flus

hed

AM

CO

burr

-up

puh

k-co

wah

; coo

ah

pert

urba

tion

(pu

hk-

cow

ah m

ale,

cooa

h fe

mal

e)y

AM

CO

hic

-up

priK

Ipo

w-u

r pe

rtur

bati

on (

pow

-ur

mal

e)y

AM

CO

hon

k

BL

RA

kic-

kic-

kerr

keek

eedr

r, de

edee

dun

kki

ckee

-doo

or

kic-

kic-

kerr

, or

ki-k

i-do

mat

e at

trac

tion

, ter

rito

rial

sig

nal

y

BL

RA

grr

krr-

krr-

krr,

grow

ling

Gro

wl,

grr-

grr-

grr,

brrr

r or

ch

urr-

chur

r-ch

urr

alar

m c

all,

terr

itor

ial d

efen

sey

BL

RA

chur

t ch

urt;

curt

; yip

, bip

or

kik;

yel

p; k

ek, k

ial

arm

cal

l

BL

RA

tch

ink-

ink-

ink

kik-

kik-

kik

or k

uk-k

uk-k

uk-k

uk; i

nk-

ink-

ink

wh

en o

n th

e n

est?

CL

RA

clat

ter

clap

per

Cla

pper

or

Cla

tter

; ch

ock-

choc

k;ca

c-ca

c-ca

c or

jupe

-jupe

-jupe

mat

e co

mm

unic

atio

ny

CL

RA

kek

ket

kek-

kek-

kek,

kik

-kik

-kik

, bup

-bup

-bup

m

ate

attr

acti

on (

by m

ale)

y

CL

RA

kek-

burr

ke

t-ket

-kar

rke

k-bu

rr

mat

e at

trac

tion

(by

fem

ale)

y

CL

RA

kek-

hur

rah

gr

unti

ng

kek-

hur

rah

y

CL

RA

hoo

H

oo; o

om-o

om-o

om

CL

RA

squa

wk

Scre

ech

or

Shri

ek; C

has

e Sq

ueal

or

kak

alar

m c

all,

terr

itor

ial d

ispu

tes

CL

RA

purr

purr

; agi

tate

d pu

rrrr

; ch

urr

CO

MO

wip

e-ou

tpe

p-pe

p-pe

hr-

peeh

rca

ckle

- ka

-ka-

ka-k

a-ka

-kee

-kre

e-kr

ee-k

ree

y

CO

MO

keep

kulp

, kee

ksq

uaw

k, y

elp,

clu

cky

CO

MO

gidd

y-up

KIR

Acl

atte

rcl

appe

rch

eup-

cheu

p-ch

eup,

jupe

-jupe

-jupe

,ge

lp-g

elp-

gelp

-gel

p-; c

hac

-ch

ac-c

hac

mat

e co

mm

unic

atio

ny

KIR

Ake

kke

tki

k-ki

k-ki

km

ate

attr

acti

on (

by m

ale)

y

KIR

Ake

k-h

urra

h

grun

tin

g

1 Nam

e(s)

of c

alls

as

liste

d in

Th

e B

irds

of N

orth

Am

eric

a (h

ttp:

//bn

a.bi

rds.

corn

ell.e

du/b

na/

).


KIR

Ake

k-bu

rrke

t-ket

-kar

rm

ate

attr

acti

on (

by fe

mal

e)y

KIR

Asq

uaw

k

LE

BI

coo

poop

oopo

o co

o or

coo

ing;

tut-t

ut-tu

t m

ate

attr

acti

ony

LE

BI

kak

rick

-ric

k-ri

ckga

ck-g

ack

mat

e co

mm

unic

atio

n, a

larm

cal

ly

LE

BI

ert

kuk

tut-t

ut-tu

t; qu

oh, h

ah o

r ca

ckle

alar

m c

all

y

LE

BI

ank-

ank

ank-

ank

wh

en fl

ush

ed

LIM

Pkr

eow

kwE

EE

eeer

, KIA

AA

arkr

eow

m

ate

attr

acti

ony

LIM

Pgo

ngo

n

PBG

Row

hoo

pge

ge

gadu

m g

adum

gw

aaaa

owse

ries

of w

ut, w

hut

or k

uk n

otes

follo

wed

by

4-20

kao

w o

r co

w n

otes

cour

tsh

ip, c

omm

unic

atio

n b

etw

een

pai

r,te

rrit

oria

ly

PBG

Rh

yen

ach

atte

rek

-ek-

ek, h

n-h

n-h

ngr

eeti

ng

call

y

PUG

Aca

ckle

pep-

pep-

pePA

A-p

ePA

A, t

o-to

-terp

Cac

kle

y

PUG

Asq

uaw

kgh

eeek

! y

PUG

Aca

c-ca

cgr

unt

Slow

Clu

ckin

g an

d G

run

t Cal

l; ca

c-ca

c-ca

c

SOR

Aw

hin

ny

wh

inn

yde

scen

din

g w

hin

ny

terr

itor

ial d

efen

se, m

ate

com

mun

icat

ion

y

SOR

Ape

r-w

eep

kooE

Epe

r-w

eep;

ker

-wee

; ter

-ee

mat

e at

trac

tion

?y

SOR

Ake

epke

ekke

e or

wee

p al

arm

cal

ly

VIR

Agr

unt

grun

tgr

unt

mat

e co

mm

unic

atio

ny

VIR

Ati

ck-it

gi

k gi

k gi

k gi

k gi

dik

gidi

k gi

dik

gidi

kti

ck-it

m

ate

attr

acti

on (

by m

ale)

y

VIR

Aki

cker

ch

i ch

i ch

i ch

i tre

errr

kick

er

solic

itat

ion

(by

fem

ale)

y

VIR

Asq

uaw

ksk

ew; k

wee

kki

ual

arm

cal

l, te

rrit

oria

l dis

pute

y

VIR

Aki

kik

kiki

k ik

-ik, p

it-ti

-ti-ti

py

YER

Acl

ick-

clic

kcl

icki

ng,

tic-

tic

tict

icti

ccl

ick-

clic

k, c

lick-

clic

k-cl

ick

mat

e at

trac

tion

y

YER

Aca

ckle

cack

leca

ckle

YER

Aw

hee

zew

hee

zin

g, c

luck

ing

wh

eeze

sh

osti

lity

App

endi

x 4.

(C

onti

nued

) L

ist o

f th

e m

ost c

omm

on c

alls

for

the

foca

l spe

cies

of

mar

sh b

irds

.

Spec

ies

Stan

dard

ized

Cal

l Nam

eSi

bley

Nam

e(s)

BN

A N

ame(

s)1

Poss

ible

Fun

ctio

nSa

mpl

e on

B

NA

1 web

site

1 Nam

e(s)

of c

alls

as

liste

d in

Th

e B

irds

of N

orth

Am

eric

a (h

ttp:

//bn

a.bi

rds.

corn

ell.e

du/b

na/

).

346 WATERBIRDS

Appendix 5. Example of a completed datasheet for a survey that followed the Standardized North American MarshBird Monitoring Protocol.

Standardized North American Marsh Bird … Standardized North American Marsh Bird Monitoring...

Documents

Transcript of Standardized North American Marsh Bird … Standardized North American Marsh Bird Monitoring...